Percentage of ICE detained people who have no criminal records
Executive summary
Current reporting from multiple research groups and news outlets converges on the conclusion that roughly seven in ten people held in ICE detention in late 2025–early 2026 had no U.S. criminal conviction on their record—most analyses place that share between about 69% and 75%—though precise figures vary by dataset and definitions used [1] [2] [3] [4].
1. What the headline numbers show: a large majority without U.S. convictions
Independent trackers and advocacy groups documented that the detained population surged in 2025 and that the composition shifted sharply toward people without criminal convictions; TRAC and the American Immigration Council reported roughly 73–75% of detainees had no criminal conviction as of late 2025, marking non‑criminal detainees as the single largest group in custody [1] [2]. The Guardian’s data reporting similarly stated that nearly 75% of the roughly 68,000 people held in mid‑December had no criminal convictions [4]. Cato Institute analyses and related writeups place comparable estimates—commonly cited figures are around 69–73% depending on whether “no criminal conviction” is measured among those arrested by ICE or among the total detained population [3] [5].
2. Why different sources give slightly different percentages
Differences in the headline percentages reflect definitional and timing issues: some counts look at the subset of people booked into custody following ICE arrests, others examine the daily average detained population, and agencies sometimes lump pending charges, immigration convictions (illegal entry/reentry), or foreign criminal records into broader categories that can skew interpretation [3] [6]. Reports also compare snapshots taken at different dates during a rapidly changing year—ICE population rose from roughly 40,000 to over 66,000 in 2025—so a November snapshot can differ materially from a mid‑December one [2] [7].
3. What “no criminal record” means in these reports—and its limits
Most public analyses cite “no criminal convictions” as recorded in U.S. criminal databases or agency classifications; that excludes pending charges, foreign convictions not surfaced in U.S. records, and immigration convictions (e.g., illegal entry) that are civil rather than criminal in many contexts, all of which can be counted differently across datasets [3] [6]. Advocates and researchers warn that ICE’s public classifications and some media summaries can blur the line between having no U.S. criminal conviction and being free of any criminal history anywhere, so the higher‑end estimates (around 73–75%) reflect U.S. conviction status, not an exhaustive global criminality measure [1] [8].
4. Context: growth driven by non‑criminal arrests and policy changes
Analysts attribute the shift largely to enforcement changes in 2025—expanded “at‑large” arrests, roving patrols, workplace raids and re‑arrests—and a surge in interior arrests of people without criminal convictions, which accounted for the overwhelming majority of detention growth that fiscal year [2] [9]. Several pieces of reporting and institutional analyses note that as detention rose—by some counts nearly 75% over the year—the net increase in detainees was overwhelmingly among those without criminal histories, with one analysis calculating that about 92% of the growth in FY2026 detention numbers came from people without criminal convictions [9] [10].
5. Bottom line, caveats, and what remains uncertain
The best available public reporting puts the share of ICE detainees with no U.S. criminal convictions in the roughly 69–75% range as of late 2025 and early 2026, but exact percentages depend on whether the denominator is nightly population, newly booked arrestees, or cumulative book‑ins, and on how pending charges, immigration convictions, and foreign records are treated [1] [3] [8]. Public agency transparency limitations and rapidly shifting enforcement practices mean that while the central claim—that a substantial majority of people in ICE custody lack U.S. criminal convictions—is well supported by multiple independent sources, precise day‑to‑day shares should be treated as estimates rather than immutable facts [2] [6].