Official Statistics

Annual Prison Performance Ratings Guide 2025/26

Published 30 July 2026

Applies to England and Wales

1. Introduction

His Majesty’s Prison and Probation Service (HMPPS) is an executive agency of the Ministry of Justice (MoJ) with the goal of helping prison and probation services work together to manage offenders through their sentences. HMPPS assesses performance for both public and privately managed prisons across England and Wales using a suite of measures that are in line with HMPPS priorities.

The Annual Prison Performance Ratings reflect performance between 1 April 2025 and 31 March 2026.

There is overlap between some information provided in the Annual Prison Performance Ratings and other MoJ publications:

• The Safety in Custody National Statistics published quarterly by MoJ. Previous and current publications can be found at https://www.gov.uk/government/collections/safety-in-custody-statistics

• The HMPPS Annual Digest Official Statistics is published annually by MoJ. Previous and current publications can be found at https://www.gov.uk/government/collections/prison-and-probation-trusts-performance-statistics

• The HMPPS Workforce Official Statistics published quarterly by MoJ. Previous and current publications can be found at https://www.gov.uk/government/collections/national-offender-management-service-workforce-statistics

• The Offender Accommodation Outcomes Official Statistics published annually by MoJ. The current publication can be found at https://www.gov.uk/government/collections/offender-accommodation-outcome-statistics

• The Offender Employment Outcomes Official Statistics published annually by MoJ. The current publication can be found at https://www.gov.uk/government/collections/offender-employment-outcome-statistics

Data presented 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.

2. Prison Performance Framework

The 2025/26 prison performance framework was used to assess the 120 HMPPS adult estate prisons. This is an increase of 1 from last year reflecting Cookham Wood which was repurposed from a YOI to an adult prison. Their targets are set based on their data as a YOI and so their final score may not yet fully reflect performance in its new role as an adult prison.

The under 18 YOIs (4) and Secure Training Centre (1) are excluded from the prison performance framework as they are monitored through the Youth Custody Service Performance Framework, summarised in the YCS Annex: 2025/26, YCS Annex Guide: 2025/26 and YCS Annex Table: 2025/26.

The 2025/26 prison performance framework used 33 outcome-focused key performance measures reflecting HMPPS priorities over the following areas:

• Safety

• Security

• Respect

• Purposeful Activity

• Preparation for Release

• Organisational Effectiveness

The majority of measures within the framework utilise a performance against target model to assess performance in each prison. See Annex A for more detail regarding the methodology, thresholds and rules applied.

The data-driven ratings were subject to a moderation process and ratified in June 2026. Overall performance in each prison is rated on a 1 to 4 scale. The different ratings are:

Figure 1: Rating definitions

Rating	Definition
4	Outstanding performance
3	Good performance
2	Performance of concern
1	Performance of serious concern

The measures by domain are shown in Figure 2 below:

Figure 2: 2025/26 Performance Measures by Priority Area^{[footnote 1]}

Domain	Measure
Safety	Prisoner on prisoner assaults incidents
	Assaults on staff incidents
	Self-harm incidents
	Positive random mandatory drug testing (rMDT)
	Number of months required MDT levels not met
	Safety Audit/Risk Management Audit
	Measuring the Quality of Prisoner Life - Safety and Security
	His Majesty’s Inspectorate of Prisons - Safety
Security	Escapes from closed prisons
	The rate of absconds
	Releases in error
	Security Audit
Respect	Racial disparity in the use of force applied in prisons
	Measuring the Quality of Prison Life - Respect
	His Majesty’s Inspectorate of Prisons - Respect
Purposeful Activity	Percentage of prisoners in at least half time purposeful activity
	Prisoners on paid work Release on Temporary License (ROTL) - rate per 1,000 prisoners
	Measuring the Quality of Prisoner Life - Purposeful Activity
	His Majesty’s Inspectorate of Prisons - Purposeful Activity
Preparation for Release	Employment at six months following custodial release
	Employment at six weeks following custodial release
	His Majesty’s Inspectorate of Prisons - Preparation for Release
	Housed on first night of custodial release
	Family and Significant Others
	Accredited Programmes completions
	Keyworker Quality Assessment
	Keyworker Quality Invalid Notes
	Offender Management in Custody Audit
Organisational Effectiveness	Incident Reporting System - Data Quality Audit
	Incident Reporting System - Assaults checks
	Incident Reporting System - Self-harm checks
	Staff resignation rate
	Staff sickness

Each performance measure along with its relevant data source can be found in Annex B.

The content of the framework is reviewed annually to ensure it remains aligned with HMPPS priorities.

The Progress made in English and maths and Prisoner attendance of education courses measures were removed from the 2025/26 framework due to most prisons migrating from the Prison Education Framework to the Prisoner Education Service part way through the performance year. The associated case management system changes meant that the data feed previously used was not comparable to the new data available. Attendance data by prison is published as an annex to this release.

The Audit of Living Conditions was also removed, due to the audit being discontinued.

The following measure has been added:

• Offender Management in Custody Audit

The following measures have been amended:

• The purposeful activity measures have been consolidated to a single measure capturing the Percentage of Prisoners in at Least Half Time Activity
• The Racial Disparity in the Use of Force measure has been revised to use more granular ethnic groupings and minimum threshold inclusion criteria
• A new version of the Keyworker Quality Assessment measure has been developed and now uses AI to score all key work case notes. This replaces the manual scoring of a sample of key work case notes.

3. Prison Scores

Each prison has an overall performance rating, which is constructed by combining all the measures in the framework.

Each measure is assigned a weighting, with the sum of all measure weights adding up to 100% for each prison. The weightings represent the importance of the performance measure for HMPPS. Measures are weighted differently depending on whether the prison is Male Closed, Female Closed, or Open, based on their relative importance. Not all performance measures in the framework apply to every prison as a result of varying prison functions.

Each applicable measure carries a rating for each prison; most are rated on a 1 to 4 scale, while some measures only allow ratings of 1 to 3. As such, the maximum possible score a prison can achieve varies depending on which measures apply to the prison and the specific distribution of weights. To overcome this, the overall prison ratings are assigned using the score the prison achieved as a percentage of the maximum possible score achievable.

There are also differences in the way measure ratings are calculated. Some compare performance to target, some are set nationally to reflect consistent expectations across the estate, whereas others are specific to each prison considering a range of local factors. For audit and HM Inspectorate of Prisons measures, the measure rating corresponds directly to the audit or inspection outcome.

The prison’s overall percentage performance score is calculated by dividing the sum of these scores multiplied by the weighting, by the sum of the maximum possible scores multiplied by the weighting. This overall percentage score determines the overall rating of the prison using the thresholds shown below in Figure 3:

Figure 3: Rating thresholds for 2025/26

Rating	Percentage of max score
4: Outstanding performance	Greater than or equal to 81%
3: Good performance	Less than 81% and greater than or equal to 58%
2: Performance of concern	Less than 58% and greater than or equal to 51%
1: Performance of serious concern	Less than 51%

4. Weights

Each measure is weighted differently for each group, reflecting the relative importance HMPPS places on the measure for that group to account for the different nature of the prisons. The agreed weightings for each group can be seen in Annex C.

4.1 Weight redistribution

Where a measure does not apply to a prison, the weighting for that measure is proportionately redistributed to all other measures across the framework, ensuring the weightings continue to sum to 100%.

Generally, Audits and HMIP Inspections do not take place on an annual basis. As such these measures are age weighted on the assumption that the results become less relevant over time. The weighting for these measures reduces in weight as shown in Figure 4 below:

Figure 4: Weighting reduction over time for Audit measures

Age	Weighting reduces by:
Up to 1 year	0%
1 to 2 years	20%
2 to 3 years	40%
3 to 4 years	60%
Over 4 years	80%

In these cases, the weighting which no longer applies to the measure is proportionately redistributed across all other measures in the framework. This includes the Measuring the Quality of Prisoner Life measures with a one year lag.

5. Rules

There are a number of rules in place across the framework that impact on the data-driven ratings for the measures. These rules can be grouped into the following categories:

• Measure level ratings being capped:

  • where, even if they meet their target, a prison cannot attain a rating above a predetermined level; or

  • performance below a certain level in an audit-based measure can enforce a cap on a non-audit based measure; or

  • where drug testing levels are not met for a certain number of months, the rating for the corresponding measure is capped.

• A measure will only apply to a prison where defined minimum thresholds have been met.

A full list of rules and the measures to which they apply can be found in Annex A.

6. Moderation

Prior to the publication of Annual Prison Performance Ratings, an annual moderation process is undertaken by HMPPS. During this process, performance data is scrutinised by key stakeholders in the prison performance process and considered alongside wider contextual evidence to make a final assessment for the year-end rating of each prison. This ensures the rating for each prison is fully reflective of performance.

In 2025/26 prisons were automatically considered for moderation if any of the following criteria was met:

• The prison had an escape in the 2025/26 performance year,

• The prison received an Urgent Notification in 2025/26 that was not considered as part of moderation in 2024/25,

• The prison received a rating of Unsatisfactory in the ‘Managing Risk of Escape’ Function of the Security Audit/Security Diagnostic in 2025/26,

• A prison in the Long Term High Security Estate (LTHSE) received a score below 3 (good) in their 2025/26 Security Audit/Security Diagnostic, or,

• The number of releases in error is above the functional group average, or the prison have been highlighted for an alternative reason by the Release in Error team.

In addition, prisons and prison leaders could submit a moderation proposal to challenge the data-driven performance rating. Where a justifiable improvement or decline in performance would move a prison to the next rating, evidence was considered by the moderation panel as to whether the rating should change.

A moderation panel discussed each proposal to agree a final rating. Prisons that were not brought forward to the moderation panel kept their data-driven rating.

In the Annual Prison Performance Ratings Publication, we present the data-driven breakdowns at measure level and whether the overall rating was changed following moderation.

7. Prison Functional Groups

Prisons were assigned to functional groups based on their predominant function. The prison functional groups were the basis of the relative performance used for assessment. Descriptions of the prison functional groups can be found in Figure 5.

Figure 5: Prison Functional Group Descriptions

Group	Description
High Security (5 prisons)	Hold male prisoners classified as category A; prisoners whose escape would be highly dangerous to the public, the police or the security of the state and for whom the aim must be to make escape impossible.
Category B (9 prisons)	Hold male prisoners classified as category B; prisoners for whom the very highest conditions of security are not necessary but for whom escape must be made very difficult.
Category C Resettlement (6 prisons)	Hold male prisoners classified as category C with up to 24 months’ time left to serve in preparation for resettlement.
Category C Trainer (15 prisons)	Hold male prisoners classified as category C; prisoners who cannot be trusted in open conditions but who do not have the will to make a determined escape attempt or the resources to do so. Some category C trainer prisons will hold a specialist cohort of men including men convicted of sexual offences. Some prisons hold foreign national prisoners who are category C and are of interest to the Home Office with between 3 and 36 months’ time left to serve of their sentence. Prisoners in training prisons may transfer to a resettlement prison with between 10 and 24 months’ time left to serve. This will be dependent on addressing their offending behaviour needs and on resettlement requirements.
Category C Trainer/Resettlement (26 prisons)	Hold male prisoners classified as category C; prisoners who cannot be trusted in open conditions but who do not have the will to make a determined escape attempt or the resources to do so. Prisoners in training/resettlement prisons will transition between the training and resettlement cohorts with between 10 and 24 months’ time left to serve in preparation for resettlement. This may mean those with 10 months’ time left to serve require a transfer to a more appropriate resettlement prison based on individual need.
Category C Foreign National (3 prisons)	Hold foreign national prisoners who are category C and are of interest to the Home Office with between 3 and 36 months’ time left to serve of their sentence.
Open (13 prisons)	Accommodate category D prisoners whose risk of absconding is considered to be low or who are of low risk to the public because of the way they have addressed their offending behaviour. Open prisons also house indeterminate and longer-sentenced prisoners who are coming towards the end of their sentence and who have gradually worked their way down the categories.
Reception (31 prisons)	Hold male prisoners. These serve the courts and receive remand and post-conviction prisoners, before their allocation to other establishments. They hold many short-term prisoners; remand prisoners; those waiting allocation to training prisons; and may hold a small number of immigration detainees. The short-term prisoners held in local prisons are those who are due for release into the surrounding area and as such engage with resettlement providers in the last 16 months of their sentence.
Female (12 prisons)	Hold female prisoners of all category classifications.

8. Comparator Groups

The 2025/26 comparator groups of prisons can be found in Table 5 of the Supplementary Tables that accompany this guide. Prisons are first grouped according to their functional group. Within these groups, a nearest neighbour statistical model is used to calculate the comparator groups. A number of contextual variables are used to determine a statistical score for each prison. These include:

• Proportion of young offenders;

• Proportion of prisoners who are sex offenders;

• Complexity of the prison to determine whether it is a standard site, complex or diverse and complex^{[footnote 2]};

• Operational capacity;

• Average age of the prisoners;

• Churn of prisoners^{[footnote 3]};

• Building age;

• Proportion of different category prisoners;

• Proportion of prisoners who are foreign national offenders;

• Urban or rural classification;

• Average sentence length of prisoners in the prison.

The variables are weighted with the most important, as set by HMPPS, contributing more to the statistically derived score. Prisons with the nearest scores are set as the comparators with each prison having a maximum of eight within their group. A comparator group is unique to the prison. If Prison A has Prisons B and C in its comparator group, this does not necessarily mean Prison B would have Prisons A and C in its group.

9. Statistical Notes

9.1 Data Redactions

1. Positive random mandatory drug testing (rMDT)

Due to reduced testing levels, reduced number of prisons with sufficient testing and the need to update the testing panel, the data is currently not sufficient to robustly estimate the percentage positive. For more information, please refer to the HMPPS Annual Digest Official Statistics. Although the data has been included in the performance calculations, due to these data quality concerns data for the Positive random mandatory drug testing is not included in Table 2.

2. Accredited programmes completions

The data for the accredited programme completions has not been finalised for 2025/26, so it is not included in the publication. This data will be included in the HMPPS Annual Offender Equalities Report, due to be released on 26 November 2026.

10. Data Quality Statement

10.1 Relevance

These statistics are generally used by:

• Government
• Academia
• Media
• Charity sector
• General public

Additionally, these statistics play an important role in helping to ensure the Ministry of Justice’s accountability to the public. To ensure these statistics meet the needs of users, users are encouraged to provide comments and feedback on existing outputs including content, breadth, frequency and methodology.

10.2 Accuracy and Reliability

These statistics are from Management Information data from administrative systems and involve manual data entry. Management Information refers to information that is not collected specifically for statistics or research. This data is collected by the Ministry of Justice for uses such as registration and record-keeping.

Although due care has been taken when processing and analysing the returns, as with any large administrative data source, the possibility of errors cannot be eliminated. While the figures shows have been checked as far as practicable, they should be regarded as approximate and not necessarily accurate to the last whole number shown in the tables.

Over time, minor changes are made to recording practices to improve the accuracy of these statistics.

In accordance with Principle 2 of the Code of Practice for Statistics, the Ministry of Justice is required to publish transparent guidance on its policy for revisions.

The 3 reasons specified for statistics needing to be revised are changes in sources of administrative systems or methodology changes, receipt of subsequent information, and errors in statistical systems and processes. Each of these points, and its specific relevance to this publication, are addressed below:

• Changes in sources of administrative systems or methodology This guidance document will clearly state where there have been revisions to data due to changes in methodology or administrative systems. In addition, statistics affected within the publication will be appropriately footnoted.

• Receipt of subsequent information The nature of any administrative system is that data may be received late. Revision mid-year will only be made where it has been deemed that the late data has a major influence on the statistics. All other minor revisions will be published in the next scheduled update.

• Errors in statistical systems and processes Occasionally errors can occur in statistical processes, however procedures are constantly reviewed to minimise this risk. Should a significant error be found, the publication on the website will be updated and an errata slip published documenting the revision.

10.3 Timeliness and Punctuality

Data in this publication is drawn from different sources, each with their own collection, validation and quality assurance processes and timelines. To account for this, data is extracted two months after the end of the performance year to allow for the most amount of time for data to mature whilst remaining practical for performance rating.

To comply with the Code of Practice for Statistics, these statistics will be announced at least 4 weeks before the publication date. Any changes to pre-announced release dates will be agreed by the Head of Profession for Statistics and promptly announced, with reasons provided for these changes.

10.4 Comparability and Coherence

The data presented in this publication was extracted in June 2026 and reflects the basis upon which prison performance was assessed. However, this may differ from other statistical publications where the data was extracted later and therefore includes any revisions. This may impact the following measures:

• For finalised data on the Prisoner on prisoner assaults incidents, Assaults on staff incidents, and Self-harm incidents measures, please refer to the Safety in Custody National Statistics publication.

• For finalised data on the Staff resignation rate measure, please refer to the HMPPS Workforce Official Statistics publication.

• For finalised data on the Housed on first night of custodial release measure, please refer to the Offender accommodation outcomes - Statistics publication.

• For finalised data on the Employment at six weeks following custodial release and Employment at six months following custodial release measures, please refer to the Offender employment outcomes - Statistics publication.

• For finalised data on the Escapes from closed prisons and Releases in error measures, please refer to the HMPPS Annual Digest publication.

10.5 Accessibility and Clarity

These statistics are published on the Ministry of Justice GOV.UK website, and the publication is available from 09:30am on the day of release.

The tables are published in an Excel format. The summary document and this guidance are both published in an accessible HTML format.

The summary document is designed in a way to provide quick access to the top-level statistics, with further information within the tables. This guidance document and the notes within the tables provide clarity on any technical terms, definitions, or acronyms to ensure that these statistics can be used effectively.

11. Annex A - Thresholds, Caps and Rules

Measure	1	2	3	4	Rules
Prisoner on prisoner assaults incidents	> 125% of target	Above target but <= 125% of target	<= Target	<= 75% of target	A national maximum rate has been included. This is the level above which a prison can only achieve a maximum rating of 2, even if they achieve their target. For 2025/26 this is set based on the function of the prison: Adult prisons: 192.11, YOI prisons (18 to 21): 457.83. Additionally, if a prison has a rating of 2 or less for the IRS Assault checks measure (or if IRS Assaults check is N/A and they have a rating of 2 or less for the IRS Data Quality Audit measure) the maximum rating achievable will be a 2.
Assaults on staff incidents	> 125% of target	Above target but <= 125% of target	<= Target	<= 75% of target	A national maximum rate has been included. This is the level above which a prison can only achieve a maximum rating of 2, even if they achieve their target. For 2025/26 this is set based on the function of the prison: Adult prisons: 84.04 YOI prisons (18 to 21): 159.53. Additionally, if a prison has a rating of 2 or less for the IRS Assault checks measure (or if IRS Assaults check is N/A and they have a rating of 2 or less for the IRS Data Quality Audit measure) the maximum rating achievable will be a 2.
Self-harm incidents	> 125% of target	Above target but <= 125% of target	<= Target	<= 75% of target	A national maximum rate has been included. This is the level above which a prison can only achieve a maximum rating of 2, even if they achieve their target. For 2025/26 the maximum rate for male prisons is 630.08 and for female prisons is 3,268.79. Additionally, if a prison has a rating of 2 or less for the IRS Self-harm checks measure (or if IRS Self-harm check is N/A and they have a rating of 2 or less for the IRS Data Quality Audit measure) the maximum rating achievable will be a 2.
Positive random mandatory drug testing (rMDT)	> 125% of target	Above target but <= 125% of target	<= Target	<= 75% of target	If a prison misses testing levels in at least one, but fewer than half of the months in the reporting period, the maximum rating achievable in the RMDT measure will be a 3. If a prison misses testing levels in half or more months in the reporting period, the rating for RMDT will be reduced by 1 and the maximum rating achievable will be a 2. If a prisons misses the testing levels in all months in the reporting period, the maximum rating achievable will be a 1. Additionally, a functional maximum rate has also been included for 2025/26. This is the level above which a prison can only achieve a maximum rating of 2, even if they achieve their target. For 2025/26 this is equal to 20.56%.
Safety Audit/Risk Management Audit	1	2	3	4
Measuring the Quality of Prison Life - Safety & Security	1	2	3	4
His Majesty’s Inspectorate of Prisons - Safety	1	2	3	4
Escapes from closed prisons	1 or more escapes	N/A	0 escapes	N/A	An escape will result in automatic moderation at the end of the year.
The rate of absconds	> 110% of target	Above target but <= 110% of target	<= Target	Top 25% prisons with lowest absconds rate on condition the target is met
Release in error
Security Audit	1	2	3	4
Racial disparity in the use of force applied in prisons					The Relative Rate Index values and associated confidence intervals are used to rate the disparity in each ethnic minority group (compared to the white population) 1-4. The lowest of these ratings is taken as the overall rating.
Measuring the Quality of Prison Life - Respect	1	2	3	4
His Majesty’s Inspectorate of Prisons - Respect	1	2	3	4
Percentage of prisoners in at least half time purposeful activity	<= 90% of target	Below target but > 90%	>= Target	> 110% of target	A prison will be awarded an automatic 4 if their outturn is 95% or more.
Prisoners on paid work ROTL - rate per 1,000 prisoners	<= 75% of target	Below target but > 75% of target	>= Target	>= 125% of target
Measuring the Quality of Prison Life - Purposeful Activity	1	2	3	4
His Majesty’s Inspectorate of Prisons - Purposeful Activity	1	2	3	4
Employment at 6 months following custodial release	<= 75% of target	Below target but > 75% of target	>= Target	> 120% of target	A national minimum percentage has been included. This is the level below which a prison can only achieve a maximum rating of 2, even if they achieve their target. For 2025/26 this is equal to 10%.
Employment at 6 weeks following custodial release	<= 75% of target	Below target but > 75% of target	>= Target	> 120% of target	A national minimum percentage has been included. This is the level below which a prison can only achieve a maximum rating of 2, even if they achieve their target. For 2025/26 this is equal to 10%.
His Majesty’s Inspectorate of Prisons - Preparation for Release	1	2	3	4
Housed on first night of custodial release	<= 90% of target	Below target but > 90% of target	>= Target	> 105% of target
Family and Significant Others	<= 90% of target	Below target but > 90%	>= Target	> 110% of target
Accredited programmes completions	<= 80% of target	Below target but > 80% of target	>= Target	N/A
Keyworker Quality	Outturn <= 2 or Invalid Cases >=5%	Outturn > 2 and Invalid Cases < 5%	Outturn > 2.5 and Invalid Cases < 2.5%	Outturn > 3 and Invalid Cases < 1%
Offender Management in Custody Audit	1	2	3	4
Incident Reporting System - Data Quality Audit	1	2	3	4
Incident Reporting System - Assaults checks
Incident Reporting System - Self-harm checks
Staff resignation rate	> 125% of target	Above target but <= 125% of target	<= Target	N/A
Staff sickness	> 12 days	Between 9 and 12 days	<= 9 days	<= 9 days and top 15% nationally

12. Annex B - Performance Measure Information

Priority Area	Performance Measure	Description	Data Source
Safety	Prisoner on prisoner assaults incidents	Prisoner on prisoner assaults incidents reported as an annualised rate per 1,000 prisoners. An assault is defined as unwanted physical contact between two or more individuals.	Digital Prison Services
	Assaults on staff incidents	Assaults on staff reported as an annualised rate per 1,000 prisoners. An assault is defined as unwanted physical contact between two or more individuals.	Digital Prison Services
	Self-harm incidents	Self-harm incidents reported as an annualised rate per 1,000 prisoners. Self-harm is defined as any act where a prisoner deliberately harms or injures themselves.	Digital Prison Services
	Positive random mandatory drug testing (rMDT)	The average rate of positive results from random mandatory drug testing.	MDT testing lab
	Number of months required MDT levels not met	The total number of months required MDT levels are not met	MDT testing lab
	Safety Audit/Risk Management Audit	An announced audit to determine how effectively the prison is assessing and managing risks of violence and self-harm. A higher score shows better performance.	Performance, Assurance and Risk Group
	Measuring the Quality of Prisoner Life - Safety and Security	A survey of prisoners’ perceptions of safety in the prison. A higher score shows better performance.	Performance, Assurance and Risk Group
	His Majesty’s Inspectorate of Prisons - Safety	Focusses on early days in custody, managing behaviour, security, safeguarding and leadership and management of safety. A higher score shows better performance.	His Majesty’s Inspectorate of Prisons
Security	Escapes from closed prisons	A prisoner escapes from a prison or prisoner escort if they unlawfully gain their liberty by breaching the secure perimeter of a closed prison, i.e. the outside wall or boundary of the prison.	Digital Prison Services
	The rate of absconds	Abscond incidents reported as a rate per 100,000 prisoner days. An abscond is an escape that does not involve overcoming a physical security. Only applies to open prisons.	Digital Prison Services
	Releases in error	A prisoner is released in error if they are released from a HMPPS establishment or court when they should otherwise have remained in HMPPS/HMCTS custody and the prisoner or third party has not deliberately played a part in the error.	Digital Prison Services
	Security Audit	To review levels of compliance within specific HM Prison Service Performance Standards.	Performance, Assurance and Risk Group
Respect	Racial disparity in the use of force applied in prisons	Assesses racial disparity in use of force by calculating the Relative Rate Index (RRI) for each ethnic minority group compared to the white population. The overall rating is based on the largest disparity identified, supported by 95% confidence intervals.	Prison NOMIS and Digital Prison Services
	Measuring the Quality of Prison Life - Respect	A survey of prisoners’ perceptions of respect in the prison. A higher score shows better performance.	Performance, Assurance and Risk Group
	His Majesty’s Inspectorate of Prisons - Respect	Focuses on staff-prisoner relationships, daily life, equality, diversity and faith, health, well-being and social care and leadership and management of respect. A higher score shows better performance.	His Majesty’s Inspectorate of Prisons
Purposeful Activity	Percentage of prisoners in at least half time purposeful activity	The percentage of prisoners in at least half-time Purposeful Activity, as a proportion of the prison population. This measure does not apply to privately managed prisons.	Prison NOMIS and Digital Prison Services
	Prisoners on paid work Release on Temporary License (ROTL) - rate per 1,000 prisoners	Prisoners released on ROTL for paid work as a rate per 1,000 prisoners in the male open estate and the fully open women’s estate.	Prison NOMIS
	Measuring the Quality of Prisoner Life - Purposeful Activity	A survey of prisoners’ perceptions of purposeful activity in the prison. A higher score shows better performance.	Performance, Assurance and Risk Group
	His Majesty’s Inspectorate of Prisons - Purposeful Activity	Focusses on time out of cell, education, skills and work activities and leadership and management of purposeful activity. A higher score shows better performance.	His Majesty’s Inspectorate of Prisons
Preparation for Release	Employment at six months following custodial release	The number of releases in employment on the date six months after release date, as a proportion of all eligible releases from a custodial sentence.	National Delius and Prison NOMIS
	Employment at six weeks following custodial release	The number of releases in employment on the date six weeks after release date, as a proportion of all eligible releases from a custodial sentence.	National Delius and Prison NOMIS
	His Majesty’s Inspectorate of Prisons - Preparation for Release	Focusses on children and families and contact with the outside world, reducing risk, rehabilitation and progression, interventions, specialist units, release planning and leadership and management of rehabilitation and release planning. A higher score shows better performance.	His Majesty’s Inspectorate of Prisons
	Housed on first night of custodial release	The number of releases with accommodation on the first night of release as a proportion of all eligible releases from a custodial sentence.	National Delius and Prison NOMIS
	Family and Significant Others	Self-assessment of services and provisions put into place within prisons to strengthen the relationship between prisoners and their family and significant others.	Family and significant other self-assessment form
	Accredited Programmes completions	The number of Offending Behaviour Programme (OBP) completions	Monthly data from prison establishments collated in central performance systems
	Keyworker Quality Assessment	An artificial intelligence assessment of the quality of key work case notes for prisoners at establishments in the male closed estate and women’s estate.	Prison NOMIS
	Keyworker Quality Invalid Notes	The percentage of case notes which do not meet the criteria of a key worker session	Prison NOMIS
	Offender Management in Custody Audit	The OMiC Key Work audit provides assurance that Key Work is being prioritised and delivered as intended in all closed adult prisons.	Performance, Assurance and Risk Group
Organisational Effectiveness	Incident Reporting System - Data Quality Audit	Audit ensuring reportable incidents are recorded on the Incident Reporting System package of Prison-NOMIS. A higher score shows better performance.	Performance, Assurance and Risk Group
	Incident Reporting System - Assaults checks	Percentage of assaults incidents checked in the Incident Reporting System Data Quality Audit recorded on Prison-NOMIS. A higher percentage shows better performance.	Performance, Assurance and Risk Group
	Incident Reporting System - Self-harm checks	Percentage of self-harm incidents checked in the Incident Reporting System Data Quality Audit recorded on Prison-NOMIS. A higher percentage shows better performance.	Performance, Assurance and Risk Group
	Staff resignation rate	Rate of Band 3-5 Prison Officer staff resignations. Some resignations are expected, but too high a rate is cause for concern. This measure does not apply to privately managed prisons.	Single Operating Platform
	Staff sickness	Average working days lost to sickness absence.	Single Operating Platform and Manual Data Return

13. Annex C - Measure Weightings

Priority Area	Measure	Weightings[footnote 4]
		Male closed	Open	Female Closed
Safety	Prisoner on prisoner assaults incidents	4.6%	3.3%	4.3%
	Assaults on staff incidents	4.6%	3.3%	4.3%
	Self-harm incidents	4.6%	3.3%	4.3%
	Positive random mandatory drug testing (rMDT)	3.1%	3.3%	2.8%
	Number of months required MDT levels not met	NA	NA	NA
	Safety Audit/Risk Management Audit	4.6%	5.0%	4.3%
	Measuring the Quality of Prisoner Life - Safety and Security	1.5%	1.7%	1.4%
	His Majesty’s Inspectorate of Prisons - Safety	4.6%	5.0%	4.3%
Security	Escapes from closed prisons	NA	NA	NA
	The rate of absconds	NA	5.0%	2.8%
	Releases in error	NA	NA	NA
	Security Audit	9.6%	6.7%	8.9%
Respect	Racial disparity in the use of force applied in prisons	3.1%	3.3%	2.8%
	Measuring the Quality of Prison Life - Respect	1.5%	1.7%	1.4%
	His Majesty’s Inspectorate of Prisons - Respect	4.6%	5.0%	4.3%
Purposeful Activity	Percentage of prisoners in at least half time purposeful activity	6.1%	5.0%	4.3%
	Prisoners on paid work Release on Temporary License (ROTL) - rate per 1,000 prisoners	1.5%	5.0%	4.3%
	Measuring the Quality of Prisoner Life - Purposeful Activity	1.5%	1.7%	1.4%
	His Majesty’s Inspectorate of Prisons - Purposeful Activity	4.6%	5.0%	4.3%
Preparation for Release	Employment at six months following custodial release	3.1%	3.3%	2.8%
	Employment at six weeks following custodial release	3.1%	3.3%	2.8%
	His Majesty’s Inspectorate of Prisons - Preparation for Release	4.6%	5.0%	4.3%
	Housed on first night of custodial release	4.6%	6.7%	4.3%
	Family and Significant Others	1.5%	5.0%	4.3%
	Accredited Programmes completions	4.6%	1.7%	4.3%
	Offender Management in Custody Audit	3.1%	NA	2.8%
	Keyworker Quality Assessment	4.6%	NA	4.3%
	Keyworker Quality Invalid Notes	NA	NA	NA
Organisational Effectiveness	Incident Reporting System - Data Quality Audit	4.6%	5.0%	4.3%
	Incident Reporting System - Assaults checks	NA	NA	NA
	Incident Reporting System - Self-harm checks	NA	NA	NA
	Staff resignation rate	3.1%	3.3%	2.8%
	Staff sickness	3.1%	3.3%	2.8%

14. Footnotes

1. Some measures are used solely in the calculation of performance scores for other measures, or to identify prisons for the end of year moderation process. These measures do not directly contribute to a prison’s overall rating. For more information on measure weightings, see Annex C. ↩

2. Prison complexity is a judgement based on prison population and churn rate, amount of staff, complexity of prisoner population and notoriety, location, site logistics, categorisation and risk, political scrutiny, media relations and management and financial commercial management. ↩

3. For the purposes of the performance framework, churn of prisoners is the rate of new admissions or transfers in as a proportion of the prison population. ↩

4. Percentages are rounded to one decimal place and therefore may not add to exactly 100%. ↩