What percentage of illegal immigrants being detained by ICE have a criminal background outside of the fact that they came into the country illegally?
Executive summary
As of late 2025 and early 2026, the best available public data show that roughly 70–75% of people in ICE custody had no U.S. criminal conviction on their records, meaning about 25–30% did have criminal convictions; several independent analyses and ICE’s own datasets converge on that ballpark [1] [2] [3]. Important caveats—differences between “arrests” and “persons in custody,” how ICE counts convictions versus pending charges, and regional variation—mean a single percentage masks meaningful complexity [4] [5] [6].
1. What the headline numbers say: major datasets point to a large non‑convicted majority in custody
Multiple trackers and analyses looking at ICE detention rosters in late 2025 find that about seven in ten people held in ICE detention lacked a U.S. criminal conviction: TRAC reported 48,377 of 65,735 detainees (73.6%) had no criminal conviction as of November 30, 2025 [1], the Cato Institute’s review of ICE data showed roughly 69% of detainees arrested by ICE had no criminal conviction by mid‑November 2025 [2], and reporting compiled by The Guardian and others put the share with no convictions at nearly 75% in December 2025 [3]. These independent sources, working from ICE’s public counts and leaked internal data, consistently place the share without criminal convictions in the roughly 69–75% range [1] [2] [3].
2. Why “arrests” and “custody” percentages diverge — definitions matter
DHS officials and some analyses stress a different metric: the share of ICE arrests with criminal charges or convictions, which can be higher than the share of people in custody with convictions, because many people arrested at the border are transferred into custody with no U.S. conviction history [5] [4]. ICE’s own materials note that detention populations include many transfers from Customs and Border Protection and that the agency groups a large “other immigration violators” category that often contains people without U.S. criminal convictions [4]. In short, “percent of arrests” and “percent of people in detention on a given day” are distinct measures and produce different headline numbers [5] [4].
3. Recent trends: growth in detention is driven largely by non‑convicted detainees
Analysts tracking changes over 2025–early 2026 found that growth in ICE daily detention counts has been concentrated among people without criminal convictions: one data analysis found that non‑criminal detainees drove roughly 72–92% of the increase in single‑day detention numbers in recent months, with only a small share of growth attributable to people with criminal convictions [7] [2]. Academic and advocacy reports also documented sharp increases in arrests and detentions of people with no recorded U.S. convictions during the same period [8] [9].
4. Local and dataset exceptions: the national average masks variation
Not every jurisdiction or dataset follows the national pattern: a state‑level ICE arrest snapshot in Utah reported that 71% of ICE arrests there involved people with criminal convictions, with just under 7% detained solely for immigration violations—illustrating how local enforcement partnerships, booking practices and data collection lead to different pictures on the ground [6]. Analysts caution that ICE arrest datasets, transfers from local jails, and Border Patrol referrals can skew local figures away from national detention‑population averages [10] [4].
5. What this does—and doesn’t—prove about public‑safety priorities
These figures demonstrate that, in late 2025, a substantial majority of people in ICE detention in the U.S. did not have recorded U.S. criminal convictions, a fact cited by advocacy groups and researchers as evidence that detention policy has expanded beyond people convicted of serious crimes [11] [9]. Proponents of current enforcement point to different metrics (arrests, pending charges, or foreign convictions not reflected in U.S. records) and note legal obligations and discretion in detention decisions; those claims are reflected in DHS and ICE statements but are not fully reconciled in the public datasets examined here [5] [4].