What official datasets (ICE/CBP/DHS) exist for custodial deaths and where can researchers access them?

Checked on January 19, 2026
Disclaimer: Factually can make mistakes. Please verify important information or breaking news. Learn more.

Executive summary

A small set of official, government-published datasets and mandated death‑report publications exist for deaths in DHS custody: ICE’s centrally posted “Detainee Death Reports,” IRS/OIG and DHS oversight reviews that aggregate ICE and CBP deaths, and DHS’s Office of Homeland Security Statistics (OHSS) operational tables that include detention events; outside projects also republish government datasets for researchers [1] [2] [3] [4]. Important caveats: ICE’s death reports are the main official line‑item source for in‑custody deaths, CBP coverage is more fragmented and largely visible via OIG reviews, and researchers commonly combine those primary sources with third‑party compilations to build longitudinal datasets [5] [2] [4].

1. ICE “Detainee Death Reports” — the primary official dataset and where to find it

ICE publishes individual “Detainee Death Reports” following a death in ICE custody and posts related news releases; statutory language in DHS appropriations requires ICE to make death reports public within a defined timeline, and ICE states it posts news releases within two business days and fuller reports within the congressional reporting window [1]. Researchers should retrieve these reports from ICE’s Detainee Death Reporting page and related ICE public affairs releases, then extract variables (name, facility, date, cause/summary) from each report for analysis [1] [5].

2. DHS Office of Inspector General (OIG) reviews — aggregated oversight datasets and analyses

The DHS OIG issues formal reviews that aggregate and analyze deaths across ICE and CBP custody—examples include a FY2021 review covering deaths in both ICE and CBP custody—which provide consolidated counts, case summaries, and systemic findings useful for cross‑agency comparison but are episodic rather than a continuously updated dataset [2]. These OIG reports are available on the DHS OIG website and are essential when researchers need validated, investigative summaries and identification of systemic policy issues [2].

3. DHS Office of Homeland Security Statistics (OHSS) operations tables — contextual detention data

DHS OHSS publishes monthly operational tables covering encounters, detention book‑ins and book‑outs, and removals; those tables are not a pure “custodial deaths” ledger but are the authoritative administrative context from which researchers can derive denominators and event timing and can link to death reports using book‑in/book‑out metadata [3] [6]. OHSS or related DHS pages and data portals are where to download those operational tables and enforcement integrated database outputs that underpin ICE administrative records [3] [6].

4. CBP deaths — limited official datapoints, mainly visible through oversight reports

CBP does not appear to maintain a single, regularly updated public “detainee death” page analogous to ICE’s death reports in the sources provided; instead, CBP deaths are documented in public statements and are the subject of OIG investigations that aggregate CBP custodial deaths for specific fiscal years [2]. Therefore, researchers seeking systematic CBP custodial‑death data will often need to rely on OIG reviews, CBP press releases, and cross‑referencing media or advocacy compilations to assemble a usable dataset [2].

5. Third‑party aggregations and how they help researchers

Academic teams and independent projects (for example, peer‑reviewed case series and the Deportation Data Project) routinely harvest ICE reports, OIG reviews, and public announcements to create analyzable datasets, noting limitations such as omitted cases (e.g., deaths after release) and differences in reporting scope between agencies [5] [7] [4]. Those secondary compilations and codebooks — available on project websites and in journal supplements — accelerate research while making transparent the provenance and gaps of government sources [4] [5].

6. Practical guidance and acknowledged limits

For a defensible dataset, researchers should start with ICE’s Detainee Death Reports (download from ICE’s site), supplement with DHS OIG reports for cross‑agency validation and systemic context, pull detention event and denominator data from OHSS operational tables, and use vetted third‑party compilations to catch reporting gaps; the available sources make clear that ICE reports deaths regardless of the place of death (facility or hospital) but that some custodial‑adjacent deaths (for example, those occurring after release) may not be captured in official “detainee death” postings without further triangulation [1] [5] [3] [2]. Where sources are silent about CBP’s own systematic public dataset, researchers must rely on OIG, press releases, and independent aggregators and be explicit about those limitations in any analysis [2] [5].

Want to dive deeper?
How complete are ICE detainee death reports compared with independent advocacy or media tallies?
What variables and metadata are consistently available in ICE Detainee Death Reports for quantitative research?
How have DHS OIG investigations of CBP custodial deaths changed reporting practices or policy recommendations?