Official Statistics

Murder, manslaughter, sexual offences and domestic abuse in the Service Justice System: 2025 Background Quality Report

Published 26 March 2026

1. Contact Details

Service Police Policy | People-CEJ-SJDC@modgovuk.onmicrosoft.com

Last updated: 26 March 2026

2. Introduction & statistical presentation

2.1 Overview

The Ministry of Defence (MOD) publishes a wide range of statistics relating to Armed Forces personnel, including statistics regarding the investigation, referral and prosecution of Persons Subject to Service Law and Civilians Subject to Service Discipline in connection with criminal conduct offences contrary to the Sexual Offences Act 2003 (SOA 03), non-recent offences, the offences of murder, manslaughter and domestic abuse.  The main purpose of these statistics is to inform policy and decision making within the Department and inform general debate within Government, Parliament and the wider public.

It is worthy of note that the Military Court Service (MCS) routinely publishes anonymised Court Martial results on the gov.uk website.

MOD does not currently produce any other Official Statistics on offending within the Armed Forces or regarding the SJS.

A transparency report relating to adult rape flagged data was published in January 2026. This was produced as an ad hoc transparency release, so is not an Official Statistic. Those numbers should not be compared to this statistic, as they count offences using different methodologies. The information can be found here SJS transparency release.

2.2 Background and Context

These statistics focus primarily on the number of offences contrary to the SOA 03 and non-recent sexual offences that were dealt with wholly within the SJS between 1 January and 31 December 2025. This includes reports to and investigated by the Service Police, all cases referred to the SPA by the Service Police or the suspect’s Commanding Officer (CO) for a prosecution decision and all charges tried by the Court Martial. The details of these cases have been used to provide statistics in respect of locations, suspects, victims, and outcomes.

Non-recent Offences are sexual offences that took place before the enactment of the SOA 03 but were reported in the reporting calendar year.

Investigations that were transferred to Civilian Police Forces or Host Nation equivalents, and those that were initially reported as relating to an offence under the SOA 03 but were later re-classified as relating to a non-SOA 03 offence have not been included in these statistics. Additionally reports regarding incidents involving other policing agencies (IIOPAs); usually where Civilian Police have jurisdiction on an incident/ investigation and inform the Service Police, are not included.

The AFA 06 requires that all charges relating to offences contrary to the SOA 03 can only be heard in the Court Martial - no sexual offences are dealt with at Summary Hearing.

As of 1 March 2018, all SOA 03 offences are listed within Schedule 2 of the AFA 06. Consequently, Section 113 of the AFA 06 requires all reports of such offences to be reported to the Service Police. All reports of a sexual offence reported to the Service Police will be investigated. All reports of murder and manslaughter are also listed within Schedule 2 of the AFA and are required to be reported to the Service Police.

Currently, decisions on which jurisdiction should deal with criminal conduct offences committed in the UK by Service personnel are made by the Service and civilian policing and prosecutorial authorities on a case-by-case basis. On 25 Oct 2023, the Director of Public Prosecutions and the Director of Service Prosecutions published a Protocol on Concurrent Jurisdiction in England and Wales, which provides guidance to prosecutors at the CPS and the SPA to assist them to decide the most appropriate jurisdiction for a particular case. A similar protocol has been published between the Director of Service Prosecutions and the Director of Public Prosecutions for Northern Ireland on 04 November 2024. A protocol with the Lord Advocate in Scotland has also been drafted and awaits final confirmation.   The Service Police are trained to undertake investigations into serious offences including murder, manslaughter and rape at the Defence School of Policing and Security. Specialist investigators also attend training accredited by the College of Policing delivered by Civilian Police Forces. Until 5 December 2022, the most serious of sexual offences and offences of murder and manslaughter were investigated by appropriately trained investigators in the Special Investigation Branch (SIB) of each respective Service Police force.  On 5 December 2022 the Special Investigation Branches of the three Services were combined to form the specialist tri-Service Defence Serious Crime Command (DSCC), under the command of Provost Marshal (Serious Crime), of which the Defence Serious Crime Unit (DSCU) is the subordinate Unit.  The Service Police (and in due course DSCU) are inspected by His Majesty’s Inspectorate of Constabulary and Fire and Rescue Services (HMICFRS). 

In March 2023, the DSCC established the Victim and Witness Care Unit (VWCU), providing support and guidance to victims of serious crimes and sexual offending.  The VWCU offers the support of a Victim Liaison Officer to all victims of sexual offending regardless of the seriousness of the case, and additionally Sexual Offence Liaison Officers are offered in circumstances when this is appropriate for the victim.

The Service Police will refer a case to the SPA once the Evidential Sufficiency Test has been met (as detailed in Section 116 of the AFA 06).

Offences contrary to the SOA 03, non-recent sexual offences and murder and manslaughter cannot be heard summarily by a CO. These must be referred by the Service Police to the SPA for a decision to prosecute or not, applying the Full Code Test contained in the Code for Crown Prosecutors (does the evidence offer a realistic prospect of conviction and is it in the public (including the Service) interest to prosecute).

The Service Prosecuting Authority (SPA) was formed on 1 January 2009. It carries out its prosecutorial functions independently of the military Chain of Command. The main role of the SPA is to review cases referred to it by the Service Police and by Commanding Officers and to prosecute appropriate cases at Court Martial. The SPA is headed by the Director of Service Prosecutions (DSP), a senior civil servant and practising criminal barrister who was appointed by HM The King.  The SPA has specialist Rape, Sexual Offence and Domestic Abuse (RASODA) teams of trained Service and civilian prosecutors.  The SPA also instructs external civilian barristers to advise and prosecute in appropriate cases.

The Military Court Service (MCS) provides a criminal court service for the Royal Navy, Army and Royal Air Force in the Court Martial. It has done so on a Tri-Service basis since 2007 and manages two permanent Military Court Centres. The Court Martial has access to the full range of special measures available in the Crown Court to assist a complainant giving evidence, and each court centre has a vulnerable witness suite with video-link facilities.  Professional witness support services are available at all courts.

Procedure in the Court Martial mirrors the civilian Crown Court in England and Wales and cases are heard by a Judge Advocate appointed by the Judge Advocate General and a Board (jury equivalent) of three or six lay members comprising officers, warrant officers and OR7 ranks.  From 1 January 2023 the law requires at least one man and one woman member (where reasonably practicable) to be specified on every Board and proceedings where five lay members were required was increased to six lay members.  Additionally, from 4 September 2023, tri-Service Boards are specified for Courts Martial.

Judge Advocates are independent civilian judges, administered by the Ministry of Justice and selected through the Judicial Appointments Commission in the same way as all civilian judges.  All Judge Advocates sit both in the Court Martial and the Crown Court. The Judge Advocate General is a Circuit Judge.

The Judge Advocate General has directed that cases involving reports of rape and sexual assault are to be listed for trial within eight weeks of the last case management direction, which means that, save in exceptional cases, trials should take place in all sexual cases within six months of the case being directed by the SPA. 

The Court Martial can impose the full range of sentences available to the Crown Court up to and including imprisonment for life, as well as a range of specific Service punishments including Service detention at the Military Corrective Training Centre.

The procedures for the Court Martial are laid down in the Armed Forces (Court Martial) Rules 2009.

3. Statistical Processing

These statistics primarily focus on the numbers of offences contrary to the SOA 03 and non-recent sexual offences that were dealt with wholly within the SJS in the 2025 calendar year, broken down into the following characteristics of interest:

• Offence type

• Country of incident

• Outcome of Service Police investigation

• Gender / Service / Status / Age of Suspect / Victim

• Cases received by SPA for prosecution decision

• Outcomes of SPA prosecution decision

• Court Martial findings by defendants / charges

Further detail is also available in the excel spread sheets that accompany these statistics.

The Service Police investigation statistics are derived from the Service Police electronic crime recording system known as CONNECT, which is used for the recording of all offences investigated by the Service Police.

Information regarding sexual offences is manually extracted from CONNECT on a weekly basis and stored on a separate database by the Crime Statistics and Analysis Cell (CSAC) within the Service Police Crime Bureau (SPCB). With regards to the data on domestic abuse, the data is extracted from the database maintained by the DSCC performance data team. This database is maintained regularly, and cases are flagged manually as domestic abuse on the database whenever a report is raised and after review by a member of the team an element of DA is identified. Where non-recent offences are presented in this publication, they are recorded on CONNECT as the equivalent offence under SOA 03, and each investigation is reviewed, and in consultation with SPA, the appropriate non-recent offence is identified. In completing this extraction and transfer of information, the content is quality controlled and verified to ensure that it is accurate as far as is possible.

The SPA case statistics are derived from the SPA electronic case recording system known as the Central Information Management Application (CIMA), which is used for the recording and tracking of all cases prosecuted by the SPA.

Information regarding sexual offences is stored on a separate database overseen by the SPA Practice Manager (Chief of Staff). In completing this extraction and transfer of information, the content is quality controlled and verified to ensure that it is accurate as far as is possible.

The MCS case statistics are derived from the MCS electronic case recording system known as the Service Courts Information Management (SCIM), which is used for the recording and tracking of all Court Martial trials managed by the MCS.

Information regarding sexual offences has been manually extracted from SCIM and cross-recorded in an in-house spreadsheet that captures the same dataset. This dual-recording process enables quality-control checks and verification to ensure the information is as accurate as far as possible.

Supplementary tables have been provided for Offences in contravention to the SOA 2003. Previous  editions did not provide these for murder/manslaughter and Indecent Images of Children due to small numbers. After review, these tables are now included to ensure as much information that is referenced in the bulletin can be found in tabulated form to increase the ease with which users can access the data.

4. Quality Management

The MOD’s quality management process for Official Statistics consists of three elements: (1) Regularly monitoring and assessing quality risk via an annual assessment; (2) Providing a mechanism for reporting and reviewing revisions/corrections to Official Statistics; (3) Ensuring BQRs are publishing alongside reports and are updated regularly.

The quality assessment of these statistics is relatively low, as the appropriate data sources are used in collaboration with the appropriate experts in each area.

Where cases in raw data extracts need clarification, for example where information may be missing from initial review of CONNECT, or where the outcome is uncertain, these records are individually examined to ascertain the correct details where possible. This is done in liaison with appropriate subject matter experts. These records are then corrected in working documents to produce the statistic, but may not match information in CONNECT as CONNECT is a live administrative data source.

5. Relevance

The MOD has previously held regular consultation meetings with users of Defence Official Statistics, which provided a forum for user feedback on their needs and perceptions. Proposed changes were set out at the consultation meetings in order to gain feedback from both internal and external users.

The MOD invites users to provide feedback to the statistical output teams on any of their publications or reports using the contact information on the front of the publication.

The principal stakeholders for these SJS statistics are the Service Police, SPA, MCS, the Military Strategic Headquarters (MSHQ) and the Conduct Equity and Justice Directorate within the Ministry of Defence (MOD). The statistics are produced by these SJS stakeholders to identify trends, develop crime prevention strategies, identify performance indicators and develop prosecution / court administrative policy and within the MOD to inform discipline, conduct and legislation policy. They are also used to answer parliamentary questions and Freedom of Information requests.

For detail on pre-release access to Defence Statistics publications please see the Defence Statistics pre-release access list webpage for the most up to date pre-release access list.

People in the roles with access receive pre-release access to the publication up to 24 hours in advance of publication and an accompanying ministerial submission.

These statistics are developed in close consultation with all SJS stakeholders.

The coverage of these statistics is close to, if not equal to all cases of offences contrary to the SOA 03 investigated by the Service Police, referred to the SPA for a prosecution decision and heard at Court Martial during the 2025 calendar year. There are no known unmet user needs.

6. Accuracy

6.1 Service Police

All offences investigated by the Service Police must be recorded on CONNECT as a matter of policy. This then generates a Unique Reference Number (URN) which allows it to be identified throughout the Service Justice System. The URN is used to manage and store all evidence in the case.

Whilst the information placed on to CONNECT is carried out by the investigator, the content is subjected to quality control by line management and the Records Office at the Service Police Crime Bureau (SPCB), or the Performance Data Team in DSCC. It is therefore as accurate as a manual system will allow.

CONNECT is a live system; consequently, the information contained within it may change as an investigation progresses. For example, an investigation into a sexual assault may change to a common assault. Interrogation of the CONNECT information for statistics may therefore produce differing results dependent on when it was conducted in relation to the change of offence. The 2025 statistics were obtained from CONNECT on 30 January 2026. Those cases under investigation on that day are therefore subject to change.

In cases of multiple offences, CONNECT will only show the result as the principal offence and this data reflects the principal offence as listed on CONNECT  If a case involves reports of rape and sexual assault, it will only show as the ‘principal offence’ and produce the result as one case of rape. Attempted offences are included in the statistics where recorded as such and where they are the principal offence. Sexual offences that are listed as a secondary offence to a non-sexual offence (i.e. not the principal offence) are also included.

6.2 Service Prosecuting Authority (SPA)

The SPA classifies each referral received as one case, regardless of the number of accused or  charges identified by the Service Police or the Commanding Officer (CO) in that referral.  The SPA statistical data reflects the number of cases referred and whether the decision following review is to bring charges in the Court Martial, to refer the case to the CO or to take no further action.

The charges which the SPA may direct be to brought in the Court Martial in respect of any particular case may not be the same as the charges which were referred for consideration to the SPA by the Service Police.. For example, a case may be referred to the SPA on suggested charges of attempted rape and sexual assault but the SPA, on reviewing the case in line with the Code for Crown Prosecutors, may decide to direct charges of assault by penetration and sexual assault. The correct charges will not be identified until all the evidence has been produced and reviewed. The Code for Crown Prosecutors involves a two stage: an evidential test, which requires prosecutors to consider whether there is a realistic prospect of making a Board (who function like a jury) sure of guilt; and public interest test. In the Service Justice System, the service interest is also taken into account: SPA lawyers are instructed that consideration of the service interest generally makes behaviour more serious. An assault or infraction that would be treated as ‘minor’ in the civilian world might have far reaching consequences for operational efficiency, unit morale or the reputation of the Armed Forces, and thus require prosecution.

CIMA is a live system, which means the information contained within it will develop as the case progresses to conclusion (e.g. direction or non-direction). Interrogation of the CIMA information for statistics may therefore produce differing results dependent on when it was conducted in relation to any change of offence. These statistics were obtained on 26 January 2026.

On receipt of the referral the case is initially recorded on CIMA and subsequently allocated to a prosecutor. This process generates an internal SPA reference number linked to the Service Police Unique Reference Number (URN). CIMA is used to record all SPA activity on every case including non-institution, referral to CO, discontinuance after direction and direction. CIMA therefore holds information on all cases considered by the SPA, although a separate database is used for statistical purposes.

The content is subjected to quality control by line management and the Practice Manager (Chief of Staff). It is therefore as accurate as a manual system will allow.

In cases where multiple suspects are reported to have jointly committed the same offence, the SPA records the matter as a single case. Consequently, some of the cases contained within the SPA statistics may have more than one accused or victim.

In cases of multiple offences, CIMA will only show the result for the most serious offence. Therefore, an investigation into a series of sexual assaults with one suspect will only produce a result as one case. Similarly, if a case involves reports of rape and sexual assault, the result will only be recorded against the most serious offence, namely rape. Offences of attempting to commit a substantive offence are included in the statistics where recorded as such and where they are the most serious offence.

6.3 Military Court Service (MCS)

All cases that are brought to trial in the Court Martial must be recorded by the MCS on SCIM on receipt of the trial direction by SPA, for the necessary pre-trial and subsequent Court Martial hearings to be arranged. This then generates an internal MCS reference number that links to the Service Police Unique Reference Number (URN) which follows the case throughout the Court Martial process. SCIM is used to record all MCS activity on each case throughout the complete Court Martial process. SCIM therefore holds information on all Court Martial trials administered by the MCS. No other organisation conducts such administration.

The information is placed on to SCIM by administrative staff on receipt of a notification that a case is required to be heard by Court Martial, following which it is updated by the MCS personnel responsible for administering the hearing each time there is a development in the case. The content is subjected to quality control by line management and the Director MCS. It is therefore as accurate as a manual system will allow.

SCIM is a live system; consequently, the information contained within it will develop and may change as a case progresses through the SPA to prosecution or not. For example, a charge of sexual assault referred for trial by Court Martial may change to an alternative non-sexual offence or no charge at all as the case proceeds to trial. Interrogation of the SCIM information for statistics may therefore produce differing results dependent on when it was conducted in relation to the change of offence. These statistics were obtained on 07 January 2026.

Defendants that were found guilty by Court Martial in 2025 but were not sentenced until 2026 are not included within these statistics.  Attempted offences are included in the statistics and are recorded within the principal offence category (for example attempted rape will be recorded within the rape offence category).

7. Timeliness and Punctuality

This is the eleventh MOD publication of statistics regarding sexual offences dealt wholly within the SJS. The final Service Police information was extracted  on 30 January 2026, in order to ensure that all cases investigated during 2025 had been uploaded on to CONNECT. The SPA information was gathered on 26 January 2026, in order to ensure that all cases referred to them during 2025 had been uploaded on to CIMA and the content confirmed as accurate. The MCS information was gathered on 07 January 2026 in order to ensure that the SCIM content was confirmed as accurate.

These statistics will continue to be published on an annual basis.

Historic and planned publication dates can be found on the UK Statistics Publication release list here: Statistics Anouncements.

The release date for this publication was pre-announced on the Calendar of upcoming statistical releases section of Statistics at MOD.

8. Coherence and Comparability

MOD Analysis Directorate published statistics on UK Armed Forces are the definitive statistics in the MOD. There are no other publicly available regular publications solely on the numbers of sexual offences dealt with by the SJS with which to ensure coherence. Within the MOD, direct queries of the CONNECT, CIMA or SCIM will produce slightly different numbers due to timing and quality issues.

These statistics may not be directly comparable with those published by other countries due to differences in the way jurisdiction of offences committed by members of the Armed Forces is managed overseas. Some countries may deal with all Armed Forces offending within their SJS, whereas others may rely more greatly on their nation’s Criminal Justice System (CJS). These statistics do not contain any cases dealt with by the CJS.

Caution must be exercised when attempting to compare conviction rates from Court Martials in the Service Justice System with those in the Crown Court in the Civilian Criminal Justice System. It is not possible to produce reliable comparisons for a number of reasons outlined below, and these statistics should not be used to do so.  

There are distinct differences between the two systems, for example the volume of cases, populations, the difference in profile, and the way the data is recorded and reported. These differences may impact the outcomes of investigations and trials. For example, differences in the rates of victim withdrawal from investigations and prosecution processes may impact the proportion of cases taken to trial, and therefore proportion of outcomes at trial. In addition, some data released by the Crown Prosecution Service reports on convictions in respect of rape-flagged cases, i.e. any case where a rape may have occurred even if it is not charged or convicted as a rape offence. This differs from this report, which provides data on the convictions for actual rape offences by defendant and by charge. 

8.1 Changes to this edition

A number of changes have been made to this edition, which fall under the four categories outlined below. There has been no removal of information available, except for the number of cases investigated and referred by service police force as this is no longer necessary due to the fact that DSCC take primacy in the cases discussed in the report. More information can be found below on specific changes.

8.2 Structural changes

• Tables have been added to count “select non sexual offences”. These are Murder, Manslaughter, Domestic Abuse and IIOC. These new or updated tables are in tabs marked as “b”. These are defined as offences that are not offences contrary to SOA 03. These numbers are not new additions to the report, but have previously either been reported in the same tab as offences contrary to SOA 03, or only in the bulletin. They have been provided in a table to aid comprehension and help users more easily use the data.

• Data from the SPA has been reformatted. Previously there were multiple tables counting each outcome as its own table, breaking it down by service police force who referred the case. Due to the change in primacy of cases, these tables have been consolidated into two tables, one showing referrals in year, and the other number of outcomes within year. See content changes below for more information.

8.3 Presentation and formatting improvements

• Columns have been re-ordered to group offences within a year, rather than by offence. This is to make it easier to understand the time period which the report relates to.

• Presentation of countries where investigations occur (excluding non-recent) has been changed to show the top 5 values, and countries associated with those values. Where there is a tie for fifth highest value, all countries are listed. At sea has also been maintained as a distinct category due to the number of offences, but is not a specific country. Due to this change, a time series is no longer provided for this table however the same information for previous years will be available in previous editions.

• Non-recent offences were previously reported with any offences which were new being added to the data table. This meant there were columns with no offences recorded for a number of years which were included in the table because there had been offences previously. To create consistency and clarity, these tables now only count Rape, Buggery, and Indecent assault (both adult and child), as they represent the largest proportion of the data. Other non-recent offences will be grouped as “other”, with details on the types of offences included in notes so as not to remove granularity of the types of non-recent offences reported.

8.4 Content changes

• A breakdown by Service Police force who own investigations or make referrals is no longer included due to the fact that Defence Serious Crime Command (DSCC) take primacy investigating serious crime in defence. Where investigation sits with single services, oversight is provided by the DSCC, however operational circumstances may apply. For this same reason, a table showing the number of non-recent investigations by service police force has been removed. The total information that would be provided by such a table is still included in the tables on the non-recent data tables so no additional information has been removed.

• Outcomes reached by the SPA (Tabs 5a and b) were previously reported as a subset of those which were referred in the year. So a case referred in 2024, that did not reach an outcome was not recorded as an outcome in the 2024 figures, but wouldn’t have been included in 2025 figures either as not referred within that year. These tables now relate to all outcomes within the year, regardless of when the case was referred to provide a clearer picture of the result of cases referred to the SPA. This has been marked with a series break to indicate that previous data is therefore not comparable.  “Referred to CPS” has also been added as an outcome to ensure all outcomes are captured.

• In the SPA outcomes table, a row has been added to show the number of cases where an outcome was reached, and a victim was under 18. This is to align with other reporting which includes breakdowns of under 18s.

• “Discontinued” has been removed as an outcome used by the Military Court Service, as it is no longer used.

9. Accessibility and Clarity

The current publication consists of a HTML report containing commentary, graphs and tables on trends in the statistics, supported by excel spreadsheets. Other formats may be possible for MOD Analysis Directorate (formerly Defence Statistics) to produce on request.

The commentary in our reports identifies and analyses the key changes in the data and provide summary statistics and policy context. Graphs, tables and other data visualisation methods are used to further explain these trends

All MOD Analysis Directorate publications can be found on the webpage: MOD Accredited and Official Statistics by topic.

They can also be accessed via the statistics release calendar on Research and statistics page of GOV.UK.

Copies of the reports are also placed in the House of Commons library.

10. Trade-offs between Output Quality Components

The main trade-off is between timeliness and quality. To ensure statistics are timely, the editing and validation process relies heavily on manual processes, which take at least a month. Spending less time investigating every relevant case would skew the accuracy of the statistics in what is an extremely sensitive field and negatively impact on the quality of our reports.

There is also a trade-off between timely reporting and waiting for cases to conclude. Statistics only include outcomes to the point where the data is taken. As a result, a number of records may be shown as awaiting outcome. These data are not routinely updated in future editions, so cases from previous years marked with no outcome may have resolved. Data is therefore representative of a snapshot in time taken at the above dates to ensure all cases have been logged for the year, in order to allow for public release of statistics as close to the year in question.

11. Performance, Cost and Respondent Burden

The Crime Statistics Analysis Cell (CSAC) within the SPCB, SPA COS and MCS administrative staff are responsible for producing information relating to investigations, prosecution and Courts Martial respectively, as well as providing analysis and advice for management information. This also includes the gathering and preparation of information in answering ad hoc enquiries and FOI requests.

There is no respondent burden as all the data is obtained from the CONNECT, CIMA and SCIM databases which are owned by each respective SJS stakeholder.

12. Confidentiality, and security

12.1 Confidentiality

All published outputs are assessed against the MOD Disclosure Control policy. MCS cases do not have disclosure control techniques applied to them as they are matters of public record, and investigations do not have disclosure control techniques applied to them as they are not directly countable against individuals.

To reduce the risk of disclosing personal information, any figure relating to a victim or suspect that is fewer than three has been suppressed. Where necessary, additional values in the same row or column have also been secondarily suppressed to prevent disclosive figures being derived from totals. Categories that already carry inherent uncertainty (such as ‘other’ or ‘unknown’) are not subject to primary suppression, to minimise unnecessary secondary suppression.

12.2 Transparency

The reports provide commentary on the key features of the outputs and identify any issues or caveats to the data. This quality report provides further information on the method, production process and quality of the output.

12.3 Security

All staff involved in the statistical production process adhere to all MOD, Civil Service and data protection regulations. The data is stored, accessed and analysed using the MOD’s restricted network and IT systems, and access to raw data is password protected and approval for access is granted only by the Department Head, Discipline, Conduct, Complaints and Legislation, Defence People Secretariat.

12.4 References

Service Prosecuting Authority.

Court Martial Results.

Statistics at MOD Homepage.

MOD Calendar of upcoming statistical releases.