Official Statistics

HMPPS Annual Digest 2024 to 2025

Published 31 July 2025

Applies to England and Wales

HMPPS Annual Digest 2024/25

Main points

Number of escapes has increased	There were 12 escapes in the 12 months to March 2025, 2 of which remained still at large 30 days after escape. This is an increase from 9 escapes the previous year, none of which remained still at large 30 days after escape.
The number of absconds and temporary release failures have decreased	In the year ending March 2025, there were 57 absconds – a 2% decrease when compared with the previous 12-month period. There were 863 temporary release failures in the year ending March 2025, of which 44 were failures to return, and 12 of these were still at large after 30 days. The number of temporary release failures decreased by 12% from the previous year.
25% of prisoners were in crowded conditions	In the 12 months to March 2025, the crowding rate at establishments across England and Wales was 25%.
The net sum earned by prisoners before the Prisoners’ Earnings Act (PEA) levy was applied decreased nominally by 1.9% compared with the same period last year	In the 12 months ending March 2025, an average of 1,080 active prisoners earned a net sum of £22.0 million before the Prisoners’ Earnings Act (PEA) levy was applied. This compares with £22.5 million earned in the previous year. A total of £3.9 million was raised through the PEA levy, down from £4.2 million last year.
rMDT is still yet to return to the levels required for reliable national or establishment-level estimates in 2024/25	In the 12 months to March 2025, there were 53,341 random mandatory drug tests (rMDT) carried out nationally across all types of drugs. This was an increase from 51,452 the previous year but is still lower than 2019- 20 when over 54,000 tests were conducted. In the 12 months to March 2025, the number of tests did not meet expected levels in any month and were below the almost 57,000 tests expected over the year.
Barricade/prevention of access incidents and incidents at height increased	In the 12 months to March 2025, there were 1,964 barricade/prevention of access incidents, an increase of 7% when compared with the previous year. There were 8,450 incidents at height in the 12 months to March 2025, an increase of 9% compared to the previous 12 months.
Finds of drugs, weapons and mobile phones increased, but alcohol has decreased	In the 12 months to March 2025 there were 26,348 incidents of drug finds, 13,014 incidents of weapons finds, and 12,166 incidents of mobile phone finds, increases of 25%, 12% and 14% respectively from the previous year. There were 8,450 incidents of alcohol finds, a decrease of 8% over the same period.
The proportion of prisoners with Enhanced (Initial) incentive status decreased	In the year to March 2025, the proportion of prisoners with an Enhanced (initial) status was 48% (an annual average of 41,537 prisoners), decreasing from the previous year when the proportion was 49% (an annual average of 42,995 prisoners).
The proportion of MBU recommendations approved increased	In the 12 months to March 2025, 90% of MBU board recommendations resulted in an approval (47 approvals out of 52 recommendations).
The number of CAS-2 referrals increased by 45% in the latest year	There were 5,308 referrals for the Community Accommodation Service in the 12 months to March 2025; an increase of 45% on the 3,666 referrals made in the previous year.
16% of HMPPS staff who declared their ethnicity were from ethnic minority backgrounds	This represents an increase of 2 percentage points compared with the previous year. HQ & Frontline Support staff had the lowest ethnic minority representation in HMPPS (14%) and staff in the Youth Custody Service (YCS) had the highest (27%).

This publication covers reporting up to and including the 2024/25 financial year. Data for the current reporting year covers the period between 1 April 2024 and 31 March 2025; and is referenced as “the 12 months to March 2025”, “the year ending March 2025” or “the latest year”.

Statisticians comment

This HMPPS Digest 2024/25 publication provides a compendium of performance data across prisons and probation for the 12 months to March 2025, with some continuing time series from previous publications, and some new and refined data:

Finds in Prison: Data on drone incidents is published in this edition for the first time.

Releases in Error: Previously, releases in error only included those that had been identified in the reporting period for those 12 months, with no revision in future years. In this publication we have additionally reported those release in error incidents that were reported after the end of the reporting period, from April 2022 onwards.

Prison Crowding: Prison crowding methodology has been refined in this publication, offering an improvement over the previous approach by using automated monthly data. This reduces the administrative burden on prisons by removing the need for manual monthly returns and provides a more reliable and accurate calculation of crowding levels across all prisons. The improvements in reporting using automated data do not materially impact the levels of crowding at a national level compared with the previous manual returns. However, the automated process has enabled corrections in reporting at prison level, resulting in some increases/decreases at a small number of prisons. For this reason, no comparisons have been made with previous years’ crowding data within this report, though similar trends can be observed: for example, the highest crowding rates continue to be seen in Reception prisons.

Introduction

His Majesty’s Prison and Probation Service (HMPPS) is an executive agency of the Ministry of Justice; with the goal of helping prison and probation services work together to manage offenders through their sentences. HMPPS replaced the National Offender Management Service (NOMS) on 1 April 2017. This HMPPS Digest 2024/25 publication provides a compendium of performance data across prisons and probation for the 12 months to March 2025.

Data presented in this report have been drawn from administrative IT systems. Although care is taken when processing and analysing the data, the level of detail collected is subject to the inaccuracies inherent in any large-scale recording system.

Products published to accompany the HMPPS Annual Digest 2024/25

The following products are published as part of this release:

• A statistical bulletin, containing commentary on key trends over time in prison performance measures and probation.
• A technical guide, providing further information on how the data are collected and processed; alongside relevant legislative or operational information relating to the topic area.
• A set of tables for each chapter (except for random mandatory drug testing) giving national and local level trends over time; and covering key topic areas in this bulletin.
• A set of data tools for specific chapters, allowing further breakdowns of the data.

Topics that are not included in this report

Information on protected characteristics of offenders is not reported here but will be published in the Offender Equalities Annual Report 2024/25 on 27 November 2025.

Data on prisoners working in custody is not included in the HMPPS Annual Digest publication due to operational changes in data collection. Prison-level data on the percentage of prisoners attending full-time and half-time purposeful activity is available in the Annual Prison Performance Ratings 2024/25 Official Statistics: https://www.gov.uk/government/collections/prison-and-probation-trusts-performance-statistics

The table on total weight of drugs found, by prison, has been removed due to the considerable limitations of this data: reporting has been shown to be variable in quality and susceptible to order effects, and because the weight of drugs is not separated by type of drug this data does not enable users to robustly gauge the number of drugs found or their impact on the prison (both density and dosage vary by drug type).

Related publications

• Offender management statistics quarterly provides detailed information on offenders held in prison custody and on probation. It includes detailed breakdowns of the prison population, prison receptions and releases. Is also covers statistics on adjudications and license recalls^{[footnote 1]}.

• Safety in custody statistics cover deaths, self-harm and assaults in prison custody in England and Wales^{[footnote 2]}.

• The HMPPS Offender Equalities Report provides protected characteristics breakdowns of placement on the incentives scheme and protected characteristic data on mother and baby units^{[footnote 3]}.

• The HMPPS Workforce Statistics provides prisons staff volumes and staff protected characteristics^{[footnote 4]}.

Review of the publication

We are continuing to review the purpose and content of the HMPPS Annual Digest. Many themes covered in this statistical release appear in related publications or could be included with material on similar topics published elsewhere. Please get in touch if you have views on the continuation of the Digest in its current form, the content, or the possible placement of related data across MoJ publications. Contact details are available on the last page of the publication.

1. Escapes, Absconds, Failure to Return from ROTL and Releases in Error

The number of escapes has increased	There were 12 escapes in the 12 months to March 2025, 2 of which remained still at large 30 days after escape. This is an increase from 9 escapes the previous year, none of which remained still at large 30 days after escape.
The number of absconds has decreased	There were 57 absconds in the year to March 2025, a 2% decrease from 58 absconds in the previous 12 months. Of the 57 prisoners who absconded, 26% (15 prisoners) remained at large for over 30 days.
The number of Release on Temporary Licence (ROTL) failures has decreased	There were 863 temporary release failures in the year to March 2025, of which, 44 were failures to return, and 12 of these were still at large after 30 days. The number of temporary release failures decreased by 12% from 985 the previous year. In the same period ROTL decreased by 10% from the previous year (from 425,095 to 381,983).
The number of releases in error has increased	There were 262 prisoners released in error in the latest year, an increase of 128% from 115 in the year ending March 2024.

The number of escapes increased from 9 to 12 in the 12 months to March 2025 (Table 1.1, Figure 1.1)^{[footnote 5][footnote 6]}

In the 12 months ending March 2025, there were a total of 12 escapes, 9 of which occurred from contractor escorts and 3 occurred from HMPPS escorts. This is an increase from 9 in the year ending March 2024. Two of the 12 escapees in the latest year remained at large 30 days after their escape.

There were no Category A prisoner escapes from prisons or HMPPS escorts. In the last 28 years (since the financial year ending March 1997), there have been two Category A escapes, occurring in the 12 months ending March 2012 and 2013.

There were no escapes from prison in the latest year. The number of escapes from prison has remained very low, not exceeding 4 in any financial year since the 12 months ending March 2005.

There were 3 escapes from HMPPS escort in the latest year. The number of escapes from HMPPS escorts has remained very low, not exceeding 4 in any financial year since the year ending March 2007.

Due to the relatively low numbers, year-on-year changes should be interpreted with caution.

Figure 1.1 shows the number of contractor escort journeys and escapes from contractor escorts. There were 557,822 contractor escort journeys in the financial year to March 2025, a 3.7% increase from 538,081 the previous year. Of the 557,822 journeys in the latest year, 9 resulted in an escape, giving a rate of approximately 1 escape per 62,000 contractor escort journeys in the latest year. The rate of escape from contractor escorts has remained consistently low at below 1 escape per 30,000 prisoner journeys each year since this series began in 2013.

Figure 1.1: Contractor escort journeys and escapes from contractor escorts, the 12 months ending March 2015 to the 12 months ending March 2025

The number of absconds decreased in the 12 months to March 2025 (Table 1.1, Figure 1.2)^{[footnote 7]}

There were 57 absconds in the year ending March 2025, a 2% decrease from 58 absconds in the previous year. The number of absconds has decreased in each of the last 5 financial years, falling from 143 in the 12 months to March 2020. Absconds had previously decreased substantially between the year ending March 2004 (1,301 absconds) and 2017 (86 absconds).

Of those who absconded in the 12 months to March 2025, 15 were still at large after 30 days. The percentage of prisoners who were at large after 30 days remained the same as the 12 months ending March 2024, at 26%.

Most abscond incidents in the 12 months to March 2025 were for prisoners whose main offence was theft (26%) or violence against the person (26%). However, prisoners whose main offence was theft or robbery had the highest rates of absconds per 1,000 prisoners in open prisons (54 absconds per 1,000 prisoners with theft offences in open prisons and 32 absconds per 1,000 prisoners with robbery offences in open prisons).

Figure 1.2: Absconds, the 12 months ending March 2004 to the 12 months ending March 2025

The number of release failures from releases on temporary licence (ROTL) decreased in the 12 months to March 2025 (Table 1.1, Figure 1.3)^{[footnote 8]}

In the year ending March 2025, there were 381,983 incidents of ROTL, a 10% decrease from 425,095 incidents in the previous year. Of those 381,983 incidents of ROTL, 863 resulted in temporary release failures, where prisoners who have been released on temporary licence fail to fulfil all of the conditions of their release. Temporary release failures decreased by 12% from 985 temporary release failures in the year ending March 2024 .

Of the 863 temporary release failures, 29 were known to have had a primary reason of committing an offence during their release, 248 of returning late back to the prison, and 44 of being failures to return, resulting in prisoners being unlawfully at large.

Of the 44 failures to return in the year ending March 2025, 12 prisoners were still at large after 30 days.

Figure 1.3: Temporary release failures from the 12 months ending March 2005 and failures to return from the 12 months ending March 2012, to the 12 months ending March 2025^{[footnote 9]}

Figure 1.3 shows that the number of temporary release failures has decreased in the latest year after an upward trend. The number of prisoners that fail to return from ROTL generally mirrored the trend for all temporary release failures prior to the pandemic, but has been broadly stable at a lower proportion of total failures in recent years. The proportion of ROTL failures relating to prisoners who fail to return from ROTL remained relatively steady at 5% in the year ending March 2025 (44 out of 863), compared to 6% in the year ending March 2024 (59 out of 985).

Most failure to return incidents in the year ending March 2025 were for prisoners whose main offence was violence against the person (30%), robbery (23 %), or drug offences (20%).

Between 2013 and 2015 a series of changes were made to tighten ROTL policy, with the result that incidences of release fell dramatically. In 2019, the MOJ issued a new ROTL policy framework, which aimed to allow prison governors to consider ROTL earlier and more frequently because of its resettlement benefits. A large-scale MOJ study published in 2018 showed that increased use of ROTL in suitable cases was associated with a small but statistically significant reduction in reoffending.

These policy changes are reflected in both incidences of release and failure. There was a decrease in the number of incidences of release on temporary licence from 530,886 in the 2013/14 financial year to 320,582 in the 2015/16 financial year, before gradually increasing to 438,117 in the 2019/20 financial year. There was a large drop to 114,856 in the 2020/21 financial year due to COVID-19 restrictions, before increasing to 301,003 in the 2021/22 financial year and 425,095 in the 2023/24 financial year. This number has dropped again to 381,983 in the most recent 2024/25 financial year. The proportion of releases completed successfully without failure remains well over 99%^{[footnote 10]}.

The number of releases in error increased in the 12 months to March 2025 (Table 1.1)

In the 12 months to March 2025, 262 prisoners were released in error. This is a 128% increase from 115 the previous year, and the highest in the time series. Of the 262 releases in error, 233 of these releases in error occurred from prison establishments, while 29 were released in error at the courts. Releases in error from establishments could also be a result of errors by the court.

Releases in error remain infrequent. The rise is believed to be linked to the requirement on Offender Management Units to digest and implement a range of operational and legislative changes^{[footnote 11][footnote 12]}. The rise in this year also partly reflects a number of offenders who were released in error in the first tranche of SDS40 due to an issue with a repealed Breach of Restraining Order offence, which was swiftly identified and corrected with legislation. These offenders were all rearrested and returned to custody.

Some releases in error incidents are only identified after the end of the reporting period for the financial year and so haven’t previously been included within the published statistics. In this publication we have additionally reported those release in error incidents that were reported after the end of the reporting period, from April 2022 onwards^{[footnote 13]}. Of the 115 incidents recorded for the year ending March 2024, 87 of those incidents were reported within the 12 month reporting period, with 28 incidents being reported after the end of the 12 month reporting period.

2. Prison Crowding

The average prison population was 87,009[footnote 14], similar to last year	In the year to March 2025, the average prison population in England and Wales was 87,009, compared with 87,129 in the previous year.
The crowding rate of prisons in England and Wales was 25%	In the 12 months to March 2025, the crowding rate at establishments across England and Wales was 25%.
Crowding rates were highest in private prisons	Crowding rates have continued to be higher in private prisons. The private prison crowding rate was 31% for the latest 12-month period, compared to 23% in public prisons.

Crowding is measured as the number of prisoners who, on the last day of the month, are held in a cell or room where the number of occupants exceeds the baseline certified normal accommodation of the cell or room. This includes the number of prisoners held two to a single cell, three prisoners in a cell designed for one or two and all prisoners held in larger cells where the occupancy exceeds the baseline certified normal capacity.^{[footnote 15]}

The maximum level of crowding for each prison is set by senior operational managers in HMPPS in agreeing the operational capacity of each establishment. No prison will be expected to operate at a level of crowding beyond that agreed by a senior operational manager.^{[footnote 16]}

Revised crowding methodology

Tables 2.2 to 2.6 are based on a refined methodology for measuring crowding. This records the proportion of the prison population held in crowded conditions and offers an improvement over the previous approach by using automated monthly data from Prison-NOMIS and Prison Cell Certificates. It identifies instances where the number of prisoners in a cell or room exceeds its Certified Normal Accommodation (CNA) - the standard for uncrowded capacity - including cases of doubling (two prisoners in a cell for one), trebling (three in a cell for one or two), and other crowded arrangements. To ensure accuracy, establishments must keep cell certificates up to date. This refined methodology reduces the administrative burden on prisons by removing the need for manual monthly returns and provides a more reliable calculation of crowding levels across all prisons.

The improvements in reporting using automated data do not materially impact the levels of crowding at a national level compared with the previous manual returns. However, the automated process has enabled corrections in reporting at prison level, resulting in some increases/decreases at a small number of prisons. For this reason, no comparisons have been made with previous years’ crowding data within this report.

The average prison population was similar to last year (Tables 2.1, 2.2 and 2.3)

In the 12 months to March 2025, the average population in prison was 87,009, a decrease of less than 1% compared with the average population of 87,129 in the previous year. In the year to March 2025, 21,616 prisoners were held in crowded accommodation conditions, which represents a crowding rate of 25%.^{[footnote 17]}

The crowding rate for public prisons is lower than the crowding rate for private prisons (Figure 2.1, Tables 2.2 and 2.4)

The crowding rate in public prisons was 23% in the year to March 2025. In privately managed prisons, the crowding rate in the latest year was 31%. The difference in crowding between public and privately managed prisons might be related to other factors, such as prison function.

Of all prisoners held in crowded conditions in the year to March 2025, 99% (21,498 prisoners) were held in doubled accommodation.^{[footnote 18]} The doubled crowding rate, as a proportion of the total prison population, was 25% in the current reporting year.

The highest crowding rates are in Reception prisons^{[footnote 19]} (Figure 2.1, Table 2.6)

Rates of crowding vary by prison function, and in the 12 months to March 2025, levels were highest in Reception prisons, where 53% (13,348) prisoners were held in crowded accommodation. Rates of crowding vary by prison function, and in the 12 months to March 2025, levels were highest in Reception prisons, where 53% (13,348) prisoners were held in crowded accommodation.^{[footnote 20][footnote 21]}

In Category C prisons, the crowding rate was 18% in the year ending March 2025. The crowding rate in Female prisons was 8% for the same time period.^{[footnote 22]}

Figure 2.1: Crowding rate in prisons across England and Wales, by prison function, 12 months ending March 2025 (Source: Table 2.6)

3. Prisoner Earnings subject to the Prisons’ Earnings Act 1996

In the 12 months to March 2025, a net sum of £22.0 million was earned by prisoners before the Prisoners’ Earnings Act (PEA) levy was applied.[footnote 23]	This is a nominal decrease of 1.9% compared with the same period last year, when £22.5 million was raised.
In the 12 months to March 2025, an average of 1,080 active prisoners were working on release on temporary licence and subject to the PEA levy.	This represents a decrease of 8.8% on the previous year, when the number was 1,183 per month.
In the 12 months to March 2025, prisoners earned an average of £1,701 per month before the levy deduction.	This represents an increase of £119 per month compared with the previous year.
In the 12 months to March 2025, a total of £3.9 million was raised through the Prisoners’ Earnings Act levy.	An average of £302 per prisoner per month was raised through the levy; totalling £3.9 million raised for the year ending March 2025. Net earnings amounted to £1,399 per prisoner per month after the levy was applied.

The Prisoners’ Earnings Act (PEA) commenced on 26 September 2011. It enables prison governors to impose a levy of up to and including 40% on net wages over £20 per week. This applies to prisoners who have been assessed as being of low risk of absconding or re-offending who are allowed to work outside of the prison on temporary licence, to prepare for their eventual release. As per the Prison Rules 1999, all monies raised from the levy are sent to Victim Support, an independent charity in England and Wales that provides specialist practical and emotional support to victims and witnesses of crime.

Total net earnings decreased from the year ending March 2024 to year ending March 2025, but average net earnings per prisoner per month before the PEA levy increased (Source: Tables 3.1 and 3.2)

In the year to March 2025, prisoners’ nominal net earnings before the levy under the Prisoners’ Earnings Act (1996) amounted to £22.0 million. This was a decrease of 1.9% (£0.4 million) from the previous year. Before the levy deduction, prisoners earned an average of £1,701 per month, £119 more per month than in the previous 12-month period. The average number of active prisoners per month was 1,080, a decrease of 8.8% from last year when the number was 1,183.

Average nominal net earnings per prisoner per month after the PEA levy increased (Source: Tables 3.1 and 3.2, Figure 3.1)

The proportion of prisoner earnings taken through the levy was 17.8% in 2024/25, a decrease from 18.7% the year before. Average nominal net earnings per prisoner per month after the levy increased, and the average amount raised for the levy per prisoner per month also increased between the year ending March 2024 and the year ending March 2025 (Figure 3.1, Table 3.2).

After the levy deduction through the Prisoners’ Earnings Act (1996), prisoners earned an average of £1,399 per month in the 12 months ending March 2025, an 8.8% nominal increase in earnings (£113) on average per prisoner per month compared with the same period in the previous reporting year (when average nominal earnings totalled £1,286 per prisoner).

Through the levy deduction, £302 per prisoner per month was raised on average for the levy. The nominal amount raised increased by 2.0% (£6) per month per prisoner in the latest year compared with the year ending March 2024, when an average of £296 was raised per prisoner per month through the Prisoners’ Earnings Act levy.

Figure 3.1: Average nominal net prisoner earnings after Prisoners’ Earnings Act levy deduction, and average raised through the levy per prisoner per month, the 6-months ending March 2012 to the 12 months ending March 2025 (Source: Table 3.2)

Total monies raised for Victim Support through the PEA levy decreased in nominal terms (Source: Table 3.1, Figure 3.2)

In the 12 months to March 2025, £3.9 million was raised for Victim Support through the levy, compared with £0.8 million in the 12 months to March 2013, the first full year for which data is available. The 2024/25 nominal amount of £3.9 million is equivalent to an inflation adjusted £2.8 million.^{[footnote 24] [footnote 25]} Figure 3.2 shows a decrease in monies raised since the previous year, following a year-on-year increase in monies raised since 2020/21 in real terms. The fall this year indicates a decrease in the amount raised for Victim Support through the levy this year, in contrast to the above inflation increases seen most years since 2012/13.^{[footnote 26]}

Figure 3.2: Total monies raised through prisoner earnings for Victim Support in nominal and real (inflation-adjusted) terms, the 6-months ending March 2012 to the 12 months ending March 2025 (Source: Table 3.1)

4. Random Mandatory Drug Testing

rMDT is still yet to return to the levels required for reliable estimates in 2024/25	In the 12 months to March 2025, there were 53,341 random mandatory drug tests (rMDT) carried out nationally across all types of drugs. This was an increase from 51,452 the previous year but is still lower than 2019-20 when over 54,000 tests were conducted. In the 12 months to March 2025, the number of tests did not meet expected levels in any month and were below the almost 57,000 tests expected over the year.

In the 12 months to March 2025, there were 53,341 random mandatory drug tests (rMDT) carried out nationally, an increase from 51,452 in 2023-24. This equates to an average of 4,445 tests per month across 111 prisons. Although the total number of tests is now close to that immediately pre-pandemic (4,500 tests per month), they remain below the required number and do not yet cover all prisons.

Due to reduced testing levels, reduced number of prisons with sufficient testing^{[footnote 27]} and the need to update the testing panel, the data is currently not sufficient to robustly estimate the percentage positive. This is also affected by the changes in which prisons are meeting their testing requirements each month. Any changes in the monthly percentage positive levels could be driven by certain prisons reaching their required levels rather than underlying levels of drug use. Since these issues affect both national estimates and prison level estimates, there are no tables accompanying this chapter of the HMPPS Digest.

A more detailed explanation of how rMDT operates is given in the Guide to the HMPPS Digest. This includes further details on why the numbers of participating prisons and testing volumes were considered too low to produce reliable and representative estimates.

Management of rMDT

In normal circumstances, as part of HMPPS’s comprehensive drug testing regime, a random sample of prisoners (5%, or 10% in prisons with under 400 prisoners) are subject to rMDT each month. This translated to over 54,000 tests completed in the year to March 2020, across all prisons. However, because of the COVID-19 pandemic, testing was suspended across prisons from April 2020 and only partially resumed from September 2020. Establishments were required to resume testing when they were operating at Stages 1 or 2 rather than at Stages 3 and 4 of the National Framework for managing COVID-19. This resulted in a significant drop in the number of completed tests throughout 2020-21, continuing into 2021-22. In April 2022 formal performance expectations around rMDT volumes were reinstated. Since then, testing levels have increased, with there being 3.7% more tests in the 12 months to March 2025, than the previous 12 months.

Testing Panel

The extent to which the testing panel covers the drugs that are prevalent in prisons, in particular the latest compounds of Psychoactive Substances (PS) in use, is another determinant of the reliability of rMDT estimates. Time lags in updating the testing panel with new substances lead to underestimation of drug use because they cannot be detected. It has not been possible to draw conclusions about the level of misuse of drugs including PS for 2018-19 and 2019-20 because of two new compounds of PS in circulation in prisons which could not at the time be identified by the rMDT test. Steps have been taken to reduce considerably the time lags in updating the rMDT testing panel for new drugs. The improvements to this process are set out in the Annex alongside the latest changes.

Because of the pause and subsequent disruption to testing due to the pandemic since 2020-21 and underestimation of drug use in 2018-19 and 2019-20 due to time lags in updating the testing panel for new PS, readers are referred to the findings in the HMPPS Digest for the period 2017-2018^{[footnote 28]}. In 2017-18, the percentage of positive drug tests (including PS) was 21%.

5. Protesting Behaviour

Barricade/prevention of access incidents increased	In the 12 months to March 2025 there were 1,964 barricade/prevention of access incidents, an increase of 7% from 1,827 in the previous 12-month period.
Hostage incidents decreased	In the 12 months to March 2025 there were 52 incidents where prisoners took someone hostage, a decrease of 10% from 58 in the previous 12 months.
Concerted indiscipline incidents decreased	In the 12 months to March 2025, the number of incidents of concerted indiscipline decreased by 24% to 220. The number of these incidents which are counted as active, incidents involving aggression or violence, decreased by 22% over the same period from 179 to 140 (64% of all concerted indiscipline incidents).
Incidents at height increased	In the 12 months to March 2025, the number of incidents at height in prisons increased by 9% to 8,450 (compared with 7,783 in the previous year).

Incidents at height continue to be the most common form of protesting behaviour (Figure 5.1, Table 5.1)^{[footnote 29]}

In the 12 months to March 2025, the most common type of incidents of protesting behaviour were incidents at height (8,450 incidents), followed by barricades or prevented access (1,964 incidents), concerted indiscipline (220 incidents) and hostages (52 incidents). Incidents of barricades or prevented access and incidents at height increased in the 12 months to March 2025 in comparison to the previous 12 months.

The increase in the latest year continued the upward trend in incidents for incidents at height and barricades or prevented access. The number of incidents at height was nearly four times higher in the 12 months to March 2025 (8,450 incidents) compared to the 12 months to March 2015 (1,726 incidents), and the number of barricades and prevented access doubled in the 12 months to March 2025 (1,964 incidents) compared to the 12 months to March 2015 (929 incidents). By comparison the number of hostage incidents decreased by 54%, from 113 incidents in the 12 months to March 2015 to 52 incidents in the 12 months to March 2025.

Figure 5.1: Number of incidents in each category of protesting behaviour, 12 months ending March 2021 to 12 months ending March 2025 (Source: Table 5.1)

The number of incidents of prisoners using barricades or preventing access increased (Source: Table 5.1)

There were 1,964 incidents in the 12 months to March 2025 where prisoners used barricades or prevented access. This represents an increase of 7% from 1,827 in the previous year, and a new peak in the time series.

Over the same period, the rate of barricade incidents per 1,000 prisoners, which takes account of the changes in the prison population, increased 8% from 21.0 per 1,000 prisoners in the previous 12 months to 22.6 per 1,000 prisoners in the 12 months to March 2025^{[footnote 30][footnote 31]}, equalling the peak of 22.6 in the 12 months to March 2020.

The number of hostage incidents decreased (Source: Table 5.1)

The number of incidents where prisoners took someone hostage in prisons decreased by 10% to 52 in the 12 months to March 2025 (compared with 58 in the previous year). This is a decrease of 60% from the peak of 129 incidents in the 12 months to March 2016.

The rate of hostage incidents in the 12 months to March 2025 decreased slightly from 0.7 to 0.6 incidents per 1,000 prisoners in comparison to the previous year. The rate remains 60% lower than the peak of 1.5 incidents per 1,000 prisoners first seen in the 12 months to March 2016.

Of the 52 hostage incidents, a large majority (42 incidents) only involved prisoners as hostages.

The number of incidents of concerted indiscipline decreased (Source: Table 5.1)

The number of incidents of concerted indiscipline decreased by 24% to 220 incidents in the 12 months to March 2025. The number of incidents decreased by 42% from the peak of 380 incidents in the 12 months to March 2017.

The number of incidents of concerted indiscipline described as active decreased by 22%, from 179 to 140 incidents in the 12 months to March 2025, while the number of incidents of described as passive decreased by 27%, from 110 to 80 incidents in the same period. The proportion of incidents of concerted indiscipline described as active increased from 62% in the previous year to 64% in the latest year, the highest percentage in the time series.

The rate of concerted indiscipline incidents per 1,000 prisoners decreased from 3.3 in the 12 months to March 2024 to 2.5 in the 12 months to March 2025, and remains 43% lower than the peak of 4.5 in the 12 months to March 2017.

The number of incidents at height increased (Source: Table 5.1)

The number of incidents at height increased by 9% to a new peak of 8,450 in the 12 months to March 2025 (compared with 7,783 in the previous year). The number of incidents at height was almost 5 times higher than the 1,726 incidents in the 12 months to March 2015.

The rate of incidents at height per 1,000 prisoners increased 9% from 89.4 in the 12 months to March 2024 to 97.1 in the 12 months to March 2025, the highest in the time series.

6. Finds in Prison

Total number of find incidents increased	In the 12 months to March 2025, the number of incidents of finds increased to 74,746 from 66,189 in the previous 12-month period, an increase of 13%.
Number of drug finds in prisons increased	In the 12 months to March 2025, the number of incidents where drugs were found in prisons increased to 26,348 from 21,145 in the previous 12-month period, an increase of 25%.
Number of mobile phone and SIM card finds increased	In the 12 months to March 2025, there were 12,166 incidents where mobile phones were found in prisons and 5,534 incidents where SIM cards were found. These are increases of 14% and 9% respectively, compared with the previous 12-month period.
Number of incidents where weapons were found increased	There were 13,014 incidents where weapons were found in the 12 months to March 2025. This is an increase of 12% compared to the previous year.
Number of tobacco finds increased	In the 12 months to March 2025, there were 5,331 incidents where tobacco was found in prisons. This is an increase of 14% compared to the previous year.
Number of alcohol finds decreased while distilling equipment finds increased	In the 12 months to March 2025, there were 8,450 incidents where alcohol was found in prisons and 671 incidents where distilling equipment was found. This is a decrease of 8% for alcohol finds but an increase of 2% for distilling equipment finds, compared with the previous 12-month period.
Number of drone incidents[footnote 32] in prisons increased[footnote 33]	In the 12 months to March 2025, the number of drone incidents in prisons increased to 1,712 from 1,196 in the previous 12-month period, an increase of 43%.

This chapter covers incidents where illicit items have been found in prisons. It is important to consider with incidents of finds in prisons, that an increase in numbers may be as a result of more items being found, although not necessarily attributable to any one particular security counter-measure, rather than more items being present in prisons.

The table on total weight of drugs found, by prison, has been removed due to the considerable limitations of this data: reporting has been shown to be variable in quality and susceptible to order effects, and because the weight of drugs is not separated by type of drug this data does not enable users to robustly gauge the number of drugs found or their impact on the prison (both density and dosage vary by drug type).

The total number of incidents where items were found increased in the 12 months to March 2025 (Table 6.1)^{[footnote 34]}

In the 12 months ending March 2025, there were 74,746 find incidents; an increase of 13% from the 66,189 in the 12 months to March 2024. Find incidents had been on an upwards trend until the pandemic, peaking at 63,764 in the 12 months to March 2020, and had then been on a downwards trend during the pandemic period until the 12 months to March 2022. Since then, find incident levels have been steadily increasing, reaching the highest level within the timeseries in the latest 12-month period.

In the year to March 2025, the most common types of illicit items found in prisons were drugs (26,348 incidents of finds); weapons (13,014); and mobile phones (12,166). There were increases in the numbers of finds of drugs, weapons, and mobile phones in the latest year, by 25%, 12%, and 14% respectively.

Figure 6.1: Number of incidents where illicit items were found in prisons, 12 months ending March 2020 to 12 months ending March 2025 (Source: Table 6.1).

The number of finds of drugs increased in the 12 months to March 2025 (Figure 6.1, Finds data tool).

In the 12 months ending March 2025, there were 26,348 finds of drugs in prisons; an increase of 25% from the 21,145 in the 12 months to March 2024. Incidents of drug finds have increased in each of the previous two years and are now at their highest level in the series. Prior to the 12 months to March 2022 they had been on a downwards trend since reaching a pre-pandemic peak of 21,575 in the 12 months to March 2020.

The number of incidents where drug equipment was found in prisons increased by 37% from 6,094 incidents in the 12 months to March 2024 to 8,325 incidents in the 12 months to March 2025.

The drug type accounting for the largest number of incidents where drugs were found in the 12 months to March 2025 were Class B drugs (Table 6.2)

There was an increase in incidents of finds where the drug type was unknown, increasing 16% to 8,445 finds, more incidents than any other drug category in this time period. The continued high volume of unknown finds is potentially due to a change in the way finds were recorded in March 2022, where a response order effect may have led to a higher proportion of ‘unknown’ responses. This may also be associated with falls in individual drug categories^{[footnote 35]}.

There was a slight decrease in finds of Class A drugs, decreasing 2% to 783 finds. Class A drugs were found in 3% of drug find incidents^{[footnote 36]}, 1 percentage point lower than the previous year. In contrast, there was a substantial increase in finds of Class B drugs, increasing by 34% to 7,705 finds, more incidents than any other drug category (beside the Unknown category) in this time period. Class B drugs were found in 29% of drug find incidents, 2 percentage points higher than the previous year.

Psychoactive substances (PS), as defined in the Psychoactive Substances Act 2016^{[footnote 37]}, were found in 6,966 incidents in the 12 months to March 2025. The number of finds of psychoactive substances had increased by 45% compared with the 12 months to March 2024, the largest increase of all drug types, and the percentage of drug finds that included psychoactive substances (PS) increased from 23% in the 12 months to March 2024 to 26% in the 12 months to March 2025.

There was a 15% increase in ‘other’ types of drugs found, to 3,532 finds.

The number of incidents where mobile phones were found increased while the number of incidents where other mobile phone related items were found decreased (Figure 6.1, and Table 6.3)

The number of incidents where mobile phones were found in prisons increased by 14% from 10,669 in the 12 months to March 2024, to a new peak of 12,166 incidents in the 12 months to March 2025. Incidents of mobile phone finds had been on an upwards trend reaching a peak of 11,792 in the 12 months to March 2020 and had since been on a downwards trend until the 12 months12 months to March 2024. The number of incidents where mobile phones were found have now overtaken pre-pandemic levels.

The quantity of mobile phones found in prisons increased by 8% over this time period, from 14,702 in the 12 months to March 2024 compared with 15,901 in the 12 months to March 2025.

The number of incidents where other mobile phone related items were found in prisons decreased 3%, from 438 in the 12 months ending March 2024 to 424 in the 12 months ending March 2025.

The number of incidents where SIM cards were found increased (Figure 6.1 and Table 6.4)

The number of incidents where SIM cards were found in prisons increased by 9% from 5,080 in the 12 months to March 2024, to a new peak of 5,534 incidents in the 12 months to March 2025.

However, the quantity of SIM cards found only increased slightly by 2.2% in this period, from 7,737 SIM cards found in the 12 months to March 2024 compared with 7,908 in the latest 12 months.

The number of incidents where memory cards were found decreased, but chargers and other digital items increased (Figure 6.1)

The number of incidents where memory cards were found in prisons decreased 13%, from 1,276 in the 12 months ending March 2024 to 1,114 in the 12 months ending March 2025.

The number of incidents where chargers were found in prisons increased 14%, from 7,925 in the 12 months ending March 2024 to a new peak of 9,064 in the 12 months ending March 2025.

The number of incidents where other digital items were found increased by 29%, from 7,010 in the 12 months ending March 2024 to 9,026 in the 12 months to March 2025.

The number of incidents where alcohol was found has decreased, while the number of incidents where distilling equipment was found increased (Figure 6.1)

The number of incidents where alcohol was found in prisons decreased by 8%, from 9,136 in the year ending March 2024 to 8,450 in the year ending March 2025.

The number of incidents where distilling equipment was found increased 2%, from 661 in the year ending March 2024 to a new peak of 671 in the year ending March 2025.

The number of incidents where tobacco was found increased (Figure 6.1)

In the 12 months ending March 2025, there were 5,331 finds of tobacco in prisons, an increase of 14% from 4,673 incidents the previous year.

The number of incidents where weapons were found increased (Figure 6.1)

The number of incidents where weapons were found in prisons increased 12% from 11,641 in the year ending March 2024 to a new peak of 13,014 in the year ending March 2025. The trend in weapon finds pre-pandemic had been increasing, peaking at 11,267 in the 12 months to March 2020. Weapon finds had a sharp drop at the start of the pandemic period, going from the peak in the 12 months to March 2020 to 7,178 in the 12 months to March 2021. Since then, weapon finds have been on an increasing trend and are now higher than pre-pandemic levels for the second year in a row.

The number of drone incidents increased (Table 6.5)

The number of drone incidents in prisons increased by 43% from 1,196 in the year ending March 2024 to a new peak of 1,712 in the year ending March 2025.

Figure 6.2: Number of drone incidents in prisons, 12 months ending March 2021 to 12 months ending March 2025 (Source: Table 6.1 and Table 6.5)

7. Incentives

The proportion of prisoners with Enhanced incentives has fallen compared with the previous year, while the proportion with Basic incentives status has risen	The proportion of prisoners with an Enhanced (initial or extended) status in the year ending March 2025 was 48%, decreasing from the previous year when the proportion was 50%. Conversely, the proportion of prisoners with a Basic status rose from 5% in 2023/24 to 6% in 2024/25.

An incentives scheme (formerly known as Incentives and Earned Privileges - IEP) was introduced in 1995 with the expectation that prisoners would earn additional privileges through demonstrating responsible behaviour and participation in work or other constructive activity. The scheme allows prisoners to earn privileges through good behaviour and engagement in the regime and rehabilitation. Privileges can also be lost through poor behaviour.

The IEP scheme was replaced by the Incentives Policy Framework, which went fully live in January 2020. The new policy has a greater focus on incentivising positive behaviour, providing consistency in key areas, whilst giving governors greater flexibility to tailor incentives to the local needs and challenges in their prison and to create levels above Enhanced.

There have been 5 incentive levels overall: Entry, Basic, Standard, Enhanced (initial) and Enhanced (extended) - although these have changed over time as levels have been added and phased out. Entry was introduced in 2013 and abolished in August 2019 while Enhanced extended began being measured from 2022.

Prisoners typically start on Standard level, and positive behaviour can be rewarded with progression through incentive levels but can also be lost by moving down an incentive level(s) for poor behaviour with the associated increase or reduction in privileges.

While the overall prison population remained similar to the previous year, there were changes in volume across each of the incentive levels. The number of prisoners on Basic status rose by 730 (17%), Standard by 468 (1%) and Enhanced (extended) by 101 (21%), while the number on Enhanced (initial) fell by 1,418 (3%) (Table 7.1).

In the year ending March 2025, an annual average of 5,001 prisoners (6%) had a Basic incentive status, an increase of 17% from the year ending March 2024 when an average of 4,272 prisoners (5%) had a Basic status. This continues the upwards trend from previous years, returning to levels seen before 2021 - before COVID-19 mitigations suspended the use of Basic incentive levels (Table 7.2, Figure 7.1).

The proportion of prisoners with a Standard incentive status remains relatively stable with a slight increase from 2023/24 (45%) to 2024/25 (46%).

The proportion of prisoners on Enhanced Initial status (48%) was lower in the year ending March 2025, compared to the previous year (49%), while those on Enhanced Extended remained stable at 1% compared to the previous year.

Figure 7.1: Percentage of Prisoners on each incentive level, in the 12 months ending March 2016 to March 2025 (Source: Table 7.2)

Youth Custody Service Young Offender Institutions had the largest proportion of prisoners with a Basic incentive status (Table 7.4)

The proportion of prisoners assigned to each type of incentive status varied greatly by prison function, in the 12 months to March 2025. The prison function with the highest proportion of prisoners on Basic incentive status was Youth Custody Service Young Offender Institutions (14% of prisoners). The prison functions with the highest proportion of prisoners on Standard incentive status were Reception and Female prisons (68% and 54% of prisoners, respectively). The prison function with the highest proportion of prisoners on Enhanced (Initial) incentive status was Open Prisons (93% of prisoners).

8. Mother and Baby Units (MBUs), Pregnant Prisoners and Births

The total number of MBU applications decreased when compared with the previous year	Between April 2024 and March 2025, 77 applications were made for a place within an MBU compared with 92 in the year to March 2024.[footnote 38] In the 12 months ending March 2025, 47 applications were approved and 5 refused.[footnote 39]
The proportion of MBU recommendations resulting in approval increased	In the 12 months to March 2025, 90% of recommendations (47 out of 52) were approved by an MBU board. In the previous year, the approval rate of recommendations was 79% (54 out of 68).
The number of women and babies received into MBUs decreased compared to last year	In the year to March 2025, 42 women and 41 babies were received into MBUs; compared with 55 women and 50 babies in the previous reporting year.
There were 196 self-declared pregnant women in prison in the 12-month period of 2024/25	Over the 12-month period of April 2024 to March 2025 there were 196 pregnant women in prison. This compares with 215 for the 2023/24 year.
There was a total of 42 births to women spending time in custody in 2024/25	All of these births (42) took place at a hospital.

A Mother and Baby Unit (MBU) is a discrete specialist accommodation unit within a women’s prison which enables mothers, where appropriate, to have their children with them. Women who are pregnant or who have children up to and around the age of 18 months can apply for a place in an MBU. Details of the process are given in the Technical Guide accompanying this publication.

There are currently six MBUs in operation across the women’s estate in England (there are no women’s prisons in Wales). These provide an overall total capacity of 64 places for mothers and 70 places for babies to allow for twins and multiple births.

This report contains statistics for self-declared pregnant prisoners and births for the financial year 2024/25. This is the third year that figures have been shown for the total number of pregnant prisoners for the year and for individual months, instead of showing weekly snapshots.

The data represents women who have self-declared as pregnant and consent to sharing this information, having been made aware why their personal data is being monitored and how it will be used. HMPPS publish these data anonymously as part of aggregate totals, preventing identification of individuals and infringement of the GDPR and Data Protection Act 2018.^{[footnote 40]} It does not represent women who have reserved their right not to disclose this personal data to HMPPS, or who might have disclosed this data to healthcare providers in confidence. Self-declared data cannot be quality assured, and therefore accuracy cannot be guaranteed.

There were 196 pregnant women in prison during the 12-month period from April 2024 to March 2025 (Table 8.2)

There were 42 births in the year April 2024 to March 2025 for women in prison, all taking place at a hospital.

The proportion of recommendations resulting in an approval for an MBU space has increased (Figure 8.1, Table 8.1)

In the 12 months ending March 2025, there were 77 applications for a place within an MBU, compared with 92 applications in the previous year. Multiple applications can be submitted by the same individual, for example if they serve two separate sentences within the time frame and have eligible children on both occasions.

Of the 52 application decisions made across the reporting year, 90% (47 applications) were approved and 10% (5 applications) refused.^{[footnote 41]} The percentage of recommendations resulting in approval during the reporting year was at the highest rate observed since the time series began in 2012, up from 79% in 2023/24 . Not all applications to MBUs will receive a decision as there are various reasons why an application may not proceed to an admissions board. For example, a woman may withdraw her application due to a change in her circumstances or because she is released on bail, or given a community sentence. Additionally, some decisions may be received in the following reporting year.

Figure 8.1: Number of applications approved and refused into an MBU, the 12 months ending March 2012 to the 12 months ending March 2025^{[footnote 42]} (Source: Table 8.1)

The number of women and babies received into MBUs decreased compared to the previous reporting year (Table 8.1)

During the latest financial year, 42 women and 41 babies were received into an MBU in England.^{[footnote 43]} This compares with 55 women and 50 babies in the 12 months ending March 2024.

The number of mothers and babies in an MBU at year end decreased compared to the previous reporting year (Table 8.1)

At the end of March 2025, 25 mothers and 22 babies were residing in an MBU in England. This compares with 38 mothers and 36 babies at the end of March 2024.

9. Community Accommodation Service (CAS-2)

The number of CAS-2 referrals increased by 45% between the 12 months ending March 2024, and the 12 months ending March 2025	There were 5,308 total referrals for the Community Accommodation Service in the 12 months to March 2025, an increase of 45% on the 3,666 referrals made in the previous year. This was driven by an increase in the number of Home Detention Curfew (HDC) referrals.
Home Detention Curfew referrals increased by 48% between the 12 months ending March 2024, and the 12 months ending March 2025	The number of HDC referrals increased by 48% in the latest year, from 2,656 referrals in the year to March 2024 to 3,922 referrals in the latest year.
Alternative to custodial recall referrals rose by 24%	The number of Alternative to custodial recall referrals increased from 111 to 138 referrals between the years ending March 2024 and March 2025.
Bail referrals increased by 15%	Referrals relating to Bail increased by 15% from 862 to 995 between the years ending March 2024 and March 2025.
CAS-2 referrals made at Category C prisons accounted for the highest volume of known prison referrals	In the year to March 2025, 45% (2,118) of all prison referrals made (where the prison name was known), were from Category C prisons. The second highest proportion were from Reception prison with 29.0% (1,359).

The Community Accommodation Service, Tier 2 (CAS-2) is a contracted service which provides short-term accommodation for those who have no suitable accommodation and may otherwise be held in custody. The service is available for those on Bail, Home Detention Curfew, on licence and at risk of recall due to loss of accommodation, and by discretion to those to be released as Homeless at Conditional Release date. The CAS-2 service provides up to 2 hours per week support to each resident; provision of CCTV in all female properties and several other properties; provision of Wi-Fi in all sites, with the issue of a basic smartphone to all residents and the rollout of wheelchair accessible bed spaces.

In the 12 months to March 2025, most CAS-2 referrals were made by prisons (88.2%) (Table 9.1, Table 9.2).

The number of CAS-2 referrals increased in the 12 months to March 2025 after a smaller rise in referrals in 2024 (Table 9.1)

In the year ending March 2025, there were 5,308 referrals to CAS-2; an increase of 45% on the 3,666 made in the year to March 2024. This follows an increased use of CAS-2 to support Bail and follows changes in the operational year to eligibility for Home Detention Curfew.

Home Detention Curfew (HDC) accounted for 74% of the CAS-2 referrals made in the year to March 2025 (Figure 9.1, Table 9.2)

HDC referrals accounted for 74% (3,922) of the total number of referrals, which is an increase in volume of 48% compared to the year ending 2024 (2,656). This represents the highest number on record, recovering from the decreases that were seen following the previous peak in 2019.

In the year to March 2025, there was also an increase in referrals relating to Alternative to Custodial Recall by 24% (from 111 to 138) and a 15% increase to Bail (court and prison) referrals (from 862 to 995 referrals).^{[footnote 44]} There was a sharp increase in referrals for Homeless at Conditional release this year, with 207 referrals compared with zero referrals last year, and eight in the 12 months ending March 2023.

Figure 9.1: Number of CAS-2 referrals by referral type, for the 12 months ending March 2019 to the 12 months ending March 2025 ^{[footnote 45]} (Source: Table 9.1)

Category C prisons had the highest volume of CAS-2 referrals (from known prisons) in the year to March 2025 (Figure 9.2, Table 9.2)

In the year ending March 2025, 45% (2,118) of CAS-2 referrals by known prisons were from Category C prisons. CAS-2 referrals made by Reception prisons accounted for 29% (1,359) of the total number of known prison referrals made in the same period.

Figure 9.2: Number of referrals made to CAS-2 by Prison Function, for the 12 months ending March 2025 (Source: Table 9.2)

10. Staff in Post

68,911 (headcount) staff in post at HMPPS	As at 31 March 2025, there were 68,911(headcount) staff in post in HMPPS, 87% of whom had declared their ethnicity.
Overall, 16% of individuals who declared their ethnicity were from an ethnic minority background	This represents an increase of 2 percentage points compared with the previous year.

Overall declaration rates have increased since 31st March 2024 (Table 10.2).

Across HMPPS overall, declaration rates have increased by 1 percentage point (from 86% as at 31 March 2024 to 87% as at 31 March 2025). Youth Custody Service staff had the highest ethnicity declaration rate (88%), representing a 1 percentage point increase since the previous year. This was the highest increase in declaration rate that was seen amongst staff. As at 31 March 2025, the declaration rate observed amongst staff in Public Sector Prisons (PSP) was 87%, and 88% amongst HQ and Frontline Support staff and there was no change from 31 March 2024. Finally, the ethnicity declaration rate amongst Probation Service staff also increased, to 85% as at 31 March 2025; an increase of 1 percentage point compared to the previous year.

HQ & Frontline Support staff have the lowest ethnic minority representation (Figure 10.1, Table 10.2)

Summary Statistics: HMPPS ethnic minority representation, and change since last year, by business area as at 31 March 2025 (Table 10.2)

	Ethnic minority representation rate	Change since 31 March 2023
Public Sector Prisons	14.6%	Increase of 2.7 percentage points
YCS	26.5%	Increase of 3.9 percentage points
HQ & Frontline Support	13.6%	Increase of 0.2 percentage point
Probation Service	18.9%	Increase of 0.9 percentage point
HMPPS Total	16.1%	Increase of 2.0 percentage points

Figure 10.1:  Ethnic minority representation rates amongst HMPPS staff, as at 31 March 2010 to 31 March 2025 (Source: HM Prison and Probation Service workforce statistics publication)

The region with the highest ethnic minority representation was London (Figure 10.2, Table 10.2)

Figure 10.2 shows the proportion of staff who are from ethnic minority backgrounds in PSP regions, the YCS and HQ & Frontline Support. Prison establishments within London had the highest representation rates with 53% of staff declaring their ethnicity as being from ethnic minority backgrounds as at 31 March 2025. In contrast, establishments in the North East had the lowest representation rate of 4%. From 31 March 2024 to 31 March 2025, all PSP regions had seen their representation rates increase and the largest increase was seen in South Central from 17% to 23%. HMPPS HQ was the only area which saw a decrease in the ethnic minority representation rate; a slight decrease of less than 1 percentage point compared to 31 March 2024.

Figure 10.3 shows the proportion of staff who are from ethnic minority backgrounds in Probation Service regions. London Probation Service had the highest representation rates with 57% of staff declaring their ethnicity as being from ethnic minority backgrounds as at 31 March 2025. In contrast, establishments in the North East Probation Service had the lowest representation rate of 4%. From 31 March 2024 to 31 March 2025 most Probation Service regions had seen their representation rates increase and the largest increases were seen in East of England Probation Service, from 15% to 17%, and Greater Manchester Probation Service, from 14% to 16%. East Midlands Probation Service and Kent, Surrey and Sussex Probation Service were the only areas which saw a decrease in the ethnic minority representation rate, of less than 1 percentage point.

Figure 10.2: Percentage of Ethnic minority staff by PSP regions^{[footnote 46]}, the YCS, HQ and Area Services, as at 31 March 2024 and 2025 (Source: Table 10.2)

Figure 10.3: Percentage of Ethnic minority (excluding white minorities) staff by Probation Service regions^{[footnote 46]}, as at 31 March 2024 and 2025 (Source: Table 10.2)

Further information

General information about the official statistics system of the UK is available from: uksa.statisticsauthority.gov.uk/

Ministry of Justice publishes data relating to offender management in England and Wales. Equivalent statistics for Scotland and Northern Ireland can be found at:
www.gov.scot/collections/crime-and-justice-statistics/

www.justice-ni.gov.uk/publications

Contact

Press enquiries should be directed as per the Ministry of Justice Press Enquiries webpage:

https://www.gov.uk/government/organisations/ministry-of-justice/about/media-enquiries

Other enquiries about these statistics should be directed to the Service Users Equality Performance Team at the Ministry of Justice:

Responsible Statistician: Danielle House

Email: SUEPer_Stats@justice.gov.uk

A copy of this report, the Technical Guide and accompanying tables can be found online:

www.gov.uk/government/collections/prison-and-probation-trusts-performance-statistics

© Crown copyright

Produced by the Ministry of Justice

Alternative formats are available on request from SUEPer_Stats@justice.gov.uk

1. Offender Management Statistics Quarterly is available at: www.gov.uk/government/collections/offender-management-statistics-quarterly ↩

2. Safety in custody statistics are available at: www.gov.uk/government/collections/safety-in-custody-statistics ↩

3. HMPPS Offender Equalities Reports are available at: www.gov.uk/government/collections/hmpps-annual-offender-equalities-report ↩

4. HMPPS Workforce Statistics are available at: www.gov.uk/government/collections/national-offender-management-service-workforce-statistics ↩

5. Unlawfully-at-large incidents are categorised by the level of security measures the prisoner had to overcome to gain their liberty. A prisoner escapes from prison if they unlawfully gain their liberty by breaching the secure perimeter of a closed prison. A prisoner escapes from an escort if they are able to pass beyond the control of escorting staff and leave the escort, the vehicle or the premises (such as a court or hospital), please see the Guide to HMPPS Annual Digest for more information. ↩

6. We report all unlawfully-at-large incidents according to the most recent reporting period. Recapture dates, which determine unlawfully-at-large incidents, are sometimes reported late and so may be recorded only after the end of the reporting period for the financial year. Therefore, unlawfully-at-large incidents included as published statistics in previous reporting periods are liable to revision. ↩

7. An abscond is an escape that does not involve overcoming a physical security barrier or restraint. It is only possible to abscond from open prison conditions. ↩

8. A temporary release failure occurs when a prisoner fails to adhere to any condition written into the licence that permits their temporary release. Failures to return after release on temporary licence are the subset of temporary release failures, where an offender has not returned to the establishment by midnight on the designated return date. ↩

9. Data on the type of temporary release failure is not available for the full time series presented in figure 1.3. Data on failures to return was first available for the 12 months ending March 2012. ↩

10. Data on the number of incidences of temporary release and individuals released, and the number of failures are reported within the Offender Management Statistics Quarterly bulletin: https://www.gov.uk/government/collections/offender-management-statistics-quarterly ↩

11. Due to the relatively low numbers, year-on-year changes should be interpreted with caution. The number of releases in error should be considered in the context of the number of releases in the same time period and changes in the operational environment. ↩

12. The Offender Management Statistics Quarterly bulletin gives an indication of the number of prisoners released in any given period via the following link: https://www.gov.uk/government/collections/offender-management-statistics-quarterly ↩

13. Some release in error incidents are only identified when a prisoner returns to custody at a later date. These incidents are captured after the end of the reporting period for the financial year, so haven’t previously been included within the published statistics. We have identified those release in error incidents that were reported after the end of the reporting period from April 2022 onwards, so figures from 2022-23 onwards may not be comparable to figures from earlier years. In this publication, we compare all release in error incidents identified within each reporting period, regardless of whether they were identified during or after the reporting period. HMPPS have invested in improving the data quality surrounding releases in error, this includes increasing capability to find incidents more quickly and also a centralised team who track and monitor incident data to identify trends and root causes that inform improvement activity which was established in 2021. ↩

14. The average prison population is based on the number of months each prison is open in the reference financial year. The annual national total is the average of each monthly total. ↩

15. For example, if 12 prisoners occupy a dormitory with an uncrowded capacity of 10, then the 12 prisoners are counted as crowded. ↩

16. The operational capacity of a prison is the total number of prisoners that an establishment can hold, considering control, security, and the proper operation of the planned regime. ↩

17. Expressed as a proportion of the total average prison population. ↩

18. Doubled accommodation is defined as two prisoners being held in a cell designed for one prisoner. ↩

19. Crowding is not evenly dispersed across the prison estate; it is particularly concentrated in Reception prisons, which are those that serve the courts of a specific area, and which predominantly hold remand and short sentenced prisoners. ↩

20. Rate of crowding is expressed as a proportion of the total prison population. The population in Reception prisons was 25,198 in the latest financial year. The total prison population across all estates in England and Wales was 87,009. ↩

21. Prison function is determined using the 2024/25 financial year list of functions. ↩

22. Crowding is less in Category B and C prisons as many of these are training prisons where activities are targeted at reducing re-offending by providing constructive regimes which address offending behaviour and improve opportunities on release. ↩

23. Net earnings are those after tax, national insurance, any court ordered payments or child maintenance payments. ↩

24. Consumer price index calculated using Bank of England inflation calculator: https://www.bankofengland.co.uk/monetary-policy/inflation/inflation-calculator ↩

25. Inflation adjusted, i.e., “real terms”, monies raised were calculated using monthly Consumer Price Index inflation figures averaged for 2012 as the base year, which is the first year for which prisoner earnings data is held. ↩

26. There was a small decrease between the 12 months ending March 2015 and 12 months ending March 2016. There was also a larger decrease between the 12 months ending March 2020 and 12 months ending March 2021, due to the impact of COVID-19. ↩

27. The maximum number of prisons with at least one test in a single month in the 12-months to March 2025 occurred in March 2025 and was 116, however only 110 of these were testing at the required levels. This is below the pre-COVID-19 average of c.120 prisons. Not testing in all prisons leads to estimates that are likely to be biased. ↩

28. The latest data including for 2017/18 and past trends are given in HMPPS Annual Digest: April 2019 to March 2020 edition (https://www.gov.uk/government/statistics/hmpps-annual-digest-april-2019-to-march2020). This includes full details of data quality which are summarised in the Guide to this year’s HMPPS Digest. ↩

29. Please see the Guide to HMPPS Annual Digest for protesting behaviour definitions. ↩

30. Rates are based on the latest prison population data published in Offender Management Statistics Bulletin within the following link https://www.gov.uk/government/collections/offender-management-statistics-quarterly. ↩

31. Please refer to the HMPPS Annual Digest Guidance Document for further information about how rates are compiled. ↩

32. Drone incidents include all drone sightings, including UAV landings, crashes, recovery, conveying unauthorised articles into the establishment restricted fly zone or where the drone is believed to be photographing prisoners, staff or parts of the buildings fabric. To be reportable as a drone sighting incident, the drone must have been seen by staff (rather than heard or observed by a third party). Where multiple drones are seen at the same time this should be recorded as one drone incident. Drone incidents are recorded separately from find incidents and are not included in the finds total. ↩

33. It is important to consider with drone incidents in prisons that an increase in numbers does not necessarily indicate more drone incursions. It may simply indicate more focused reporting. Trends should not be directly attributed to any one particular security counter-measure. ↩

34. There was a change in the way Finds were recorded in March 2022, please see the Guide to HMPPS Annual Digest for more information. ↩

35. There was a change in the way Finds were recorded in March 2022, please see the Guide to HMPPS Annual Digest for more information. ↩

36. Figures on the percentage of incidents by class of drugs are approximations that assume no find incidents include multiple drugs within each class. ↩

37. In this Act “psychoactive substance” means any substance which is capable of producing a psychoactive effect in a person who consumes it, and is not an exempted substance. Exempted substances are: controlled drugs, medicinal products, alcohol or alcoholic products, nicotine, tobacco products, caffeine or caffeine products or any substance which is ordinarily consumed as food, and does not contain a prohibited ingredient. http://www.legislation.gov.uk/ukpga/2016/2/contents ↩

38. Applications received in one year may be approved or refused in a following year. ↩

39. The remaining applications did not proceed to an admissions board during this reporting period either because the application was withdrawn for e.g. because they were released from custody or a decision had not been reached by the time of reporting. ↩

40. Where figures are so low that they risk identifying individuals, we have marked them as less than 3 to avoid identification. ↩

41. An application approval does not always mean a woman and her baby/babies will be received into an MBU, as circumstances can change after a decision is made. For example, the mother may be released, or a legal care order could be obtained by Children’s Services whereby the Local Authority can override decisions about the child’s placement. ↩

42. The percentage of approved applications is the proportion of those who received a recommendation in the year (of approvals or refusals). ↩

43. The number of women received into an MBU does not necessarily equal the number of approvals within a financial year period. This is because an application for a woman can be approved in the previous financial year, but she does not enter the MBU until the current financial year. ↩

44. Alternative to Custody Recall and Homeless at Conditional Release replaced Additional Licence Conditions from 1st October 2022, under the new CAS-2 Nacro contract. ↩

45. Figure 11.1 includes only the two referral types with the largest volume of referrals, with referrals for Additional Licence Conditions, Alternative to Custodial Recall, Homeless at Conditional Release, Intensive Alternative to Custody, Intensive Supervision Court, Move on from Approved Premises and Risk Assessed Recall Review grouped under “Other”. ↩

46. Information reflects the latest regional management structure implemented. These structures do not necessarily reflect geographical border areas. ↩ ↩²