CATCH: The Pentagon's Quiet Database That Waits for a S...

A service member who has just been sexually assaulted faces a choice that almost no one outside the military hears about. She can file what the Department of Defense calls a Restricted Report, which unlocks a forensic exam, medical care, and a victim advocate while keeping her name out of any criminal investigation, or she can file an Unrestricted Report, which starts an investigation and, with it, the exposure and career risk that keep most victims silent in the first place. For years the restricted path bought privacy at a hidden cost. If the person who assaulted her had done the same thing to someone else, in a different barracks or a different year, nobody would ever connect the two accounts. The Catch a Serial Offender Program, known inside the services as CATCH, exists to close exactly that gap. It is a small, anonymous database run by the Naval Criminal Investigative Service, and the honest way to measure it is not in dollars, which the Pentagon never breaks out, but in submissions. There were 2,999 of them by September 30, 2024, producing 155 suspect matches.

What CATCH is and why it exists

CATCH is a voluntary, anonymous database that lets an adult sexual assault victim who has filed a Restricted Report submit details about the suspect and the incident so that investigators can check whether the same person has been named by anyone else. The victim does not lose the confidentiality of restricted reporting by using it. She fills out the CATCH portion of the reporting paperwork, the entry goes into a restricted server, and nothing about her identity is exposed unless she later chooses to make it so. If, and only if, someone else has independently named the same suspect, both victims are privately notified and each is offered the choice to convert her Restricted Report into an Unrestricted Report and open an investigation. No one is forced. The match is an invitation, not a subpoena.

The reason the program exists is a hard fact about how these crimes work. Most military sexual assaults are never reported at all, and a disproportionate share of the ones that do surface are committed by people who offend more than once. Restricted reporting, which the department created to coax more victims into care, had an unintended side effect. It made repeat offenders harder to see, because each restricted account sat in isolation. CATCH is the department's answer to that trade-off. It tries to preserve the privacy that gets victims to come forward while still building a quiet ledger of who has been accused, so that a pattern can announce itself even when no single victim is ready to press a case.

The program went fully operational on August 5, 2019. In July 2024 the department widened who can use it. A Deputy Secretary of Defense memorandum that month, titled "Updates to Department of Defense Policies to Enhance Support for Adult Sexual Assault Victims," extended CATCH eligibility to some service members and adult dependents who never filed an official report, through a separate entry form, and to DoD civilians in cases where the suspect was a service member. That expansion is recent enough that its effect on the numbers is still working its way into the annual reports.

The mechanism and the people who run it

CATCH is a Department of Defense program, but it is operated at the headquarters level by one service on behalf of all of them. The Naval Criminal Investigative Service runs the centralized, restricted server and the searchable database for every branch and the National Guard. Specially designated investigators from NCIS, Army Criminal Investigation Division, and Air Force Office of Special Investigations do the comparison work. They check each new entry against every other entry and against open and closed investigative files, looking for the same name, the same description, the same fact pattern. When two entries point to the same suspect, the notification and conversion offer go back out to the victims through the same channels that keep the process confidential.

The designation of who runs this is not informal. It traces to a specific issuance. DoD Instruction 5505.18, "Investigation of Adult Sexual Assault in the Department of Defense," is the governing document, effective March 22, 2017, with a later change effective July 26, 2024. Its paragraph on CATCH states that, per a Deputy Secretary of Defense memorandum dated December 28, 2016, the Secretary of the Navy was designated as the Executive Agent for the program. That December 2016 memorandum, listed in the instruction's references, specifically names the Navy as Executive Agent for the development and maintenance of the Catch a Serial Offender program server and searchable database. The DoD Executive Agent registry carries the same designation. So the line of authority is clean. The Navy owns it, NCIS operates the server, and investigators from three services do the matching.

One design detail matters more than it looks. CATCH entries are retained for up to ten years. A victim who submits an entry, then transfers to another command or leaves the service entirely, can still be reached by a later match years down the line. The program is built to be patient. It assumes that the second victim who unlocks a pattern may not come forward for a long time, and it keeps the door open until then.

The money, measured in submissions

There is no separate dollar figure for CATCH, and any honest accounting has to say so plainly. It is not a major acquisition, not a weapons platform, not a contract with a ceiling. It is a modest information system living inside NCIS and inside the much larger Sexual Assault Prevention and Response budget, and the department does not publish a line item that isolates its cost. So the scale has to be read in submissions, and there the primary numbers are specific.

The FY2020 Annual Report on Sexual Assault in the Military reported 444 submissions and 11 matches for fiscal year 2020 alone, as of September 30, 2020, and then added a cumulative line: since its launch in August 2019, CATCH had received 636 victim submissions resulting in 25 matches. That cumulative figure, 636 and 25, is a since-launch total, not a single-year count, and it is worth keeping the two straight because they are easy to blur. A companion report on the service academies added roughly 43 midshipman and cadet submissions in academic year 2020 to 2021, though that academy sub-figure rests on a secondary source rather than a primary document I could confirm, so it should be treated as approximate.

The trail since then runs in one direction, upward. As of September 30, 2021, the cumulative total was 915 submissions and 35 matches. A year later it was 1,614 and 63. A year after that, 2,309 and 109. The most recent figure, from the FY2024 Annual Report on Sexual Assault in the Military, is 2,999 total victim submissions producing 155 matches as of September 30, 2024, with 690 submissions and 46 matches in fiscal year 2024 alone. Those are the load-bearing numbers for anyone weighing the program, and they are drawn from the department's own primary reporting rather than from an outside estimate.

The honest efficiency critique

The weakness is not fraud or a runaway budget. It is thin yield. Across every year of data, matches run at roughly five percent of submissions. In the FY2020 cumulative total it was 25 out of 636. By FY2024 it was 155 out of 2,999. The ratio is remarkably stable, which tells you the low hit rate is structural, not a startup wobble that got fixed. For every twenty victims who take the trouble to file a CATCH entry, only about one finds a corroborating account waiting.

There is a deeper limit. A match, on its own, changes nothing. It is not an investigation, not a charge, not an accountability action. It only matters if a victim then chooses to convert her Restricted Report to an Unrestricted one, and the program cannot require that. Some victims, offered the match, still decline to open a case, which is their right and often their safest choice, but it means the pipeline from match to prosecution is narrow. The department does not publish a clean count of how many CATCH matches actually produced investigations and convictions, and I did not find a dedicated Government Accountability Office or Defense Advisory Committee study scoring the program's cost effectiveness. So this critique leans on the program's own reported numbers rather than an independent audit, and that gap is itself part of the honest picture. A watchdog has not yet put CATCH on a scale. It is also fair to note that low usage may reflect the underlying reluctance to report at all, which no database can fix, rather than a flaw in the tool.

The honest public-good defense

For the narrow problem it was built to solve, CATCH is well designed and humane. Start from the two facts that justify it. Most military sexual assaults are never reported, and repeat offenders account for a disproportionate share of the ones that are. Any mechanism that can surface a hidden link between two isolated accounts, at low risk to the people who provide them, is doing something genuinely useful that nothing else in the system does. CATCH preserves the confidentiality that draws victims into medical care and advocacy in the first place, and it hands them a victim-controlled path to accountability that opens only if and when someone else names the same person.

The low match rate cuts both ways here. Five percent sounds small until you translate it into people. By September 2024 the program had identified 155 matches, and each match represents a repeat-offender link that would otherwise have stayed invisible, a second victim who now knows she is not the only one, and a decision point that did not exist before. It does this cheaply, without a major acquisition, and without forcing anyone into an investigation before she is ready. Measured against its actual purpose, which is to see patterns that the privacy of restricted reporting would otherwise hide, the program works.

Reading the ledger

Both readings are true at once, and the fair thing is to let them stand together rather than collapse them into a headline. On the efficiency side, CATCH is a small program with a low hit rate, a narrow path from match to accountability, and no outside audit of whether it earns its keep. On the public-good side, it fills a real hole in a system built around victim privacy, it costs little, it forces no one, and it has surfaced 155 repeat-offender links that the old approach would have buried. The department's own primary reporting supports both statements without contradiction. This is not a boondoggle in the usual sense of money poured into a hole. If anything it is the opposite risk, a quiet and underused tool that may be doing more good per dollar than almost anything else in the sexual assault portfolio, precisely because so few people know it exists. The ledger reading is that CATCH is cheap, humane, and structurally limited, and that the strongest criticism of it is not that it wastes money but that no one has yet measured it, and too few victims have been told it is there.

Fact-check notes and sources

The Navy (Secretary of the Navy, with NCIS operating the server) is the designated DoD Executive Agent for the CATCH program server and searchable database, per a Deputy Secretary of Defense memorandum dated December 28, 2016, and the governing issuance is DoD Instruction 5505.18, "Investigation of Adult Sexual Assault in the Department of Defense" (effective March 22, 2017, with a change effective July 26, 2024). DoD Executive Agent registry and DoDI 5505.18.
CATCH is a voluntary, anonymous program run by NCIS for all services and the National Guard that lets a Restricted Report victim learn whether the same suspect was named by another victim; it became fully operational on August 5, 2019, entries are retained for up to ten years, and eligibility was expanded in July 2024 to some non-reporting service members, adult dependents, and DoD civilians. DoD SAPR CATCH program page, U.S. Army launch announcement, and the NCIS CATCH victim entry portal.
The 636 submissions and 25 matches figure is a cumulative since-launch total as of the FY2020 report, not a single fiscal year count; the FY2020-only figures were 444 submissions and 11 matches as of September 30, 2020. The most recent cumulative figure is 2,999 total victim submissions and 155 matches as of September 30, 2024, from the FY2024 Annual Report on Sexual Assault in the Military. The match-to-submission ratio holds near five percent across every year. Psychology Today summary citing the DoD annual reports and the DoD annual reports at sapr.mil.
The roughly 43 service academy submissions for academic year 2020 to 2021 are attributed to the DoD report on sexual harassment and violence at the military service academies but rest on a secondary source rather than a primary document I could independently confirm, so that sub-figure is approximate and should be treated as such. Psychology Today.
The department does not publish a separate dollar figure for CATCH; it is a modest information system inside NCIS and inside the wider Sexual Assault Prevention and Response budget, so the program's cost is not broken out publicly and its scale is best read in submissions rather than spending. DoD SAPR CATCH program page.

This post is informational and journalistic, drawn from public records, and is not legal, financial, or policy advice; all figures are attributed to their fiscal year as reported by the Department of Defense.

CATCH: The Pentagon's Quiet Database That Waits for a Serial Offender to Repeat