Keep Factually independent
Whether you agree or disagree with our analysis, these conversations matter for democracy. We don't take money from political groups - even a $5 donation helps us keep it that way.
How many of the people detained by ICE have a criminal record?
Executive Summary
ICE detention data show a substantial portion of people held have no criminal convictions: multiple datasets report between 65% and 71.5% of detainees lacked criminal convictions in mid-to-late 2025. ICE and some state analyses emphasize different measures and time frames — for example, ICE has at times reported a higher share of arrestees with convictions or pending charges — so the difference largely reflects how “criminal record” is defined and which population (booked, current detainees, or arrestees) is counted [1] [2] [3].
1. A surprising majority: recent snapshots that flip the conventional story
Three independent analyses published in 2025 converge on a striking finding: a clear majority of people in ICE custody had no criminal convictions at the time of booking or while in detention. Nonpublic ICE booking data and follow-up reporting indicate 65% of people booked into detention had no convictions and over 93% had no violent convictions, framing most detainees as nonviolent [1]. A separate public “Immigration Detention Quick Facts” snapshot reported 71.5% of current detainees had no criminal convictions as of September 21, 2025, listing 42,755 of 59,762 detainees without convictions [2]. These figures challenge the common perception that ICE detention primarily holds people with criminal histories, and they highlight the importance of distinguishing conviction status from arrest or charge status [1] [2].
2. ICE’s framing: convictions, charges, and selective time windows
ICE’s own enforcement metrics present a different angle: agency releases and operational statistics have at times stated that roughly 70% of individuals it arrested had convictions or pending charges, a figure that mixes convictions and unresolved charges and focuses on arrests rather than the detained population at a point in time [3]. That framing inflates the share with criminal involvement by counting pending charges or by selecting a subset of enforcement actions (for example, interior arrest operations), rather than the full detained population. Because policy debates hinge on whether detainees are “criminals,” ICE’s emphasis on convictions-plus-pending-charges reflects an operational and political agenda to justify enforcement priorities; independent datasets using booking records and conviction histories yield markedly lower shares of convictions [3] [1].
3. State and local snapshots complicate the national picture
State-level analyses show variation that matters. The Texas Tribune’s November 3, 2025, analysis finds that in Texas a higher share of ICE arrestees under the Biden administration had criminal convictions (58%) than under the prior administration, illustrating how geography, local enforcement partnerships, and time period change the mix of people ICE detains [4]. Local enforcement priorities, 287(g) partnerships, and the selection of operational targets (e.g., workplace raids versus border arrests) produce divergent mixes of detainees. These regional differences demonstrate that national aggregates mask substantial heterogeneity, so claims about “how many detainees are criminals” require specifying location and which ICE action is measured [4].
4. Definitions and methodological choices drive the headline numbers
The gap between 65–71.5% with no convictions and ICE’s ~70% with convictions or charges is largely methodological. Public dashboards, point-in-time detainee counts, booking records, and arrest logs each capture different populations; counts may include pending charges, traffic offenses, immigration violations, or prior deportation orders that are not criminal convictions [2] [1] [3]. Reporting that lumps minor infractions, traffic violations, or immigration-related offenses with serious criminal convictions inflates the perceived criminality of detainees. Independent analyses explicitly separating convictions, pending charges, and offense severity show that violent convictions are rare among those detained [1].
5. What the data leave out and why that matters for policy
Existing datasets often omit contextual details crucial for policy debates: time since last conviction, whether convictions are state versus federal, plea outcomes, or whether charges stem from immigration-related conduct. ICE dashboards provide demographic and population trends but do not always make conviction-level granularity publicly accessible for independent verification [5] [6]. The nonpublic ICE booking data made available to reporters allowed deeper analysis showing low violent-offense rates, but reliance on such limited-access data means the public record can shift when new extracts become available. For policymakers, the distinction between past convictions, pending charges, and noncriminal administrative grounds for detention should guide decisions about detention use, alternatives, and resource allocation [1] [5].
6. Bottom line: the numbers disagree — because they measure different things
Recent, credible analyses from mid- to late-2025 consistently show that a majority of people in ICE custody at points in time lacked criminal convictions (65–71.5%), with most convictions nonviolent or minor [1] [2] [7]. ICE’s alternate messaging — that a large share of its arrestees have convictions or pending charges — reflects a different measurement choice that mixes arrests, charges, and selective enforcement windows [3]. The correct interpretation depends on the question: if you ask “who is currently detained?” the best available public and journalistic analyses point to most detainees having no criminal convictions; if you ask “who has ICE arrested in specific operations?” the answer can show a higher criminality rate.