What percentage of illegal aliens arrested by ICE have criminal records?
Executive summary
There is no single authoritative percentage because it depends entirely on which metric and dataset are used: the Department of Homeland Security (DHS) publicly asserts that roughly 70% of ICE arrests involve people "convicted or charged" with crimes in the United States [1] [2], while multiple independent analyses and leaked ICE data show that a large share of people arrested or detained by ICE have no U.S. criminal convictions — estimates in recent reporting range from about 41% to more than 70% with no convictions depending on timing and definitions [3] [4] [5] [6].
1. DHS’s claim: 70% are “criminal” when charged or convicted
DHS and ICE messaging in January 2026 repeatedly framed enforcement results as targeting the "worst of the worst," stating that roughly 70% of ICE arrests are of “criminal illegal aliens” who have been convicted or charged in the United States, a figure used across multiple DHS press releases and briefings [1] [2] [7]. DHS’s language typically bundles together people with convictions and people with pending charges, and it highlights individual high‑profile arrests—murderers, sex offenders and gang members—to reinforce that aggregate figure [8] [9].
2. Independent and watchdog data: a very different picture on convictions
Leaked ICE datasets and third‑party analyses show a far lower share of detainees with U.S. criminal convictions: by late July 2025, the Cato Institute’s review of internal ICE arrest data found about 67% of arrests were of people without criminal convictions and only about 5% had violent convictions [6]. TRAC’s public data tracking ICE detention reported that, as of November 30, 2025, roughly 73.6% of people held in ICE detention had no criminal conviction on record, and TRAC also reported the raw number of people booked into detention during October 2025 [5]. Other analyses of ICE’s shifting detained population note that the share of detainees with no criminal record rose sharply through 2025, with some reporting that around 41–48% of detainees lacked a criminal history at different points in late 2025 and early 2026 [3] [4].
3. Why the numbers diverge: definitions, data slices and transfers matter
These conflicting figures stem from differences in definitions (convictions vs. charges vs. “criminal alien” classifications), timing and scope: DHS often counts anyone with charges or pending cases while watchdogs emphasize proven convictions; many ICE detainees are transfers from Border Patrol and thus typically lack U.S. convictions even if removable by statute; and datasets vary over fiscal periods and between arrests, bookings and people in custody on a given snapshot date [10] [11] [5]. ICE itself notes that custody determinations consider criminal history among many other factors, and that arrests include both administrative immigration violations and criminal warrant enforcement [11] [12].
4. What the best summary looks like given available reporting
If the question is strictly “what percentage of illegal aliens arrested by ICE have criminal records (convictions)?” recent independent reporting supports the conclusion that a majority of people ICE has arrested or detained in 2025–early 2026 did not have U.S. criminal convictions — figures cited include roughly 67%–74% with no convictions in some datasets [6] [5], but point‑in‑time snapshots from ICE and advocacy groups produce lower no‑conviction shares (about 41–48%) depending on the date and whether CBP transfers are included [3] [4]. If the question includes people with pending charges, DHS’s public claim of about 70% falls within the range DHS itself promotes but is not corroborated by independent data that separates charges from convictions [1] [2] [6].
5. Caveats, bias and what remains unanswered
Confidence is limited by opaque, shifting datasets and competing institutional agendas: DHS and ICE emphasize high‑profile criminal removals to justify expanded enforcement [1] [8], while academic and advocacy groups highlight leaked internal data and booking snapshots that show large numbers of non‑convicted detainees and rapid growth in interior arrests [6] [3] [5]; PolitiFact and other fact‑checkers have engaged these claims but the full methodological reconciliation of DHS’s 70% figure with independent datasets is not present in the materials provided [13]. Given these limits, the most defensible answer is conditional: DHS claims ~70% when counting convictions plus charges [1] [2], while independent tracking of convictions alone shows that a substantial majority of those arrested or detained by ICE at multiple points in 2025–2026 had no U.S. criminal conviction — often in the 60–74% range depending on the dataset and date [6] [5] [4].