What share of ICE detainees are held for immigration offenses (e.g., illegal entry/reentry) versus state/federal criminal convictions?
This fact-check may be outdated. Consider refreshing it to get the most current information.
Executive summary
As of the fall of 2025, multiple independent data compilations and reporting show that roughly two-thirds to three-quarters of people in ICE custody had no U.S. criminal conviction, with those convicted of U.S. crimes making up roughly a quarter to a third of the detained population and the remainder holding pending charges or being detained solely for immigration violations [1] [2] [3]. ICE’s own public statistics categorize detainees into those with convictions, those with pending criminal charges, and those with no convictions or pending charges but removable for immigration law violations — a framing that helps explain variation across snapshots and analyses [4].
1. The headline numbers: a clear majority of detainees lack U.S. criminal convictions
Multiple independent trackers and news organizations analyzing ICE’s public data found that about 70–74% of people in ICE custody at key points in 2025 had no U.S. criminal conviction; for example, TRAC reported roughly 72% no conviction in September and TRAC/other compilations put the no-conviction share near 73.6% as of late November [5] [1]. CBS reported that on Nov. 16, 2025 nearly half of detainees — about 48% — lacked any U.S. criminal charges or convictions and were held solely on civil immigration violations, while ICE-listed convicted detainees represented about 26% of the total on that date [2].
2. Pending charges and immigration-only arrests complicate the picture
Beyond the simple binary of “convicted” versus “not convicted,” a meaningful share of detainees have pending criminal charges; national snapshots show a nontrivial portion of custody is people with pending charges rather than finalized convictions, and many who have been detained by ICE were picked up from local jails before criminal cases resolved [4] [6]. Reporting by Stateline and Prison Policy Project underscores that ICE relies heavily on state and local arrests and jail transfers — situations where the ultimate criminal outcome (dismissal, acquittal, conviction) may not be captured in a single ICE snapshot [7] [6].
3. Growth since 2025 skewed heavily toward non‑criminal detainees
Analysts tracking fiscal‑year and calendar changes found that most of the recent surge in ICE detention has been among people without U.S. criminal convictions: one analysis estimated over 90% of FY2026 detention growth was driven by detainees with no convictions, and others report that non‑criminal detainees accounted for the vast majority of increases since early 2025 [8] [9]. TRAC and Migration Policy noted that although the detained population rose sharply, the share with criminal convictions fell, sometimes by half, during key periods in 2025 [9] [3].
4. The administration’s public claims versus the audited data
DHS and ICE communications have repeatedly emphasized that enforcement prioritizes “the worst of the worst,” with statements claiming 70% of ICE arrests involve people with convictions or pending charges; independent data researchers and multiple news outlets have challenged that portrayal, pointing to the agency’s own tables and TRAC figures that show a larger share lacking U.S. convictions [10] [5]. ICE’s public press releases catalog high‑profile arrests of people convicted of serious crimes, which are factual but do not, by themselves, change the larger statistical distribution captured in the agency’s detention snapshots [11].
5. What the data do — and do not — resolve
ICE’s statistics and third‑party analyses reliably indicate that a substantial majority of detainees at multiple checkpoints in 2025 were being held for immigration violations rather than U.S. criminal convictions, while a smaller but significant minority had convictions and another slice had pending charges [4] [1] [2]. Available public sources do not always disaggregate the severity of criminal convictions (violent vs. minor) uniformly, nor do they consistently capture foreign convictions, outstanding warrants, or classified national‑security flags that ICE cites to justify some arrests, so precise attribution across those categories remains limited in public reporting [2] [4].