What percentage of arrests by ICE are criminals
This fact-check may be outdated. Consider refreshing it to get the most current information.
Executive summary
ICE detention and arrest data from 2025 show a persistent pattern: a large majority of people in ICE custody have no criminal conviction — reported figures cluster around roughly 71–74% with no conviction in detention (e.g., 71% reported by Migration Policy and 73.6% reported by TRAC) [1] [2]. The administration and DHS present competing figures and framing, claiming their arrests largely target people with criminal convictions (DHS asserted 70% of arrests were of those “convicted or charged”), while multiple independent analyses and news outlets show the opposite picture for detainees overall [3] [4] [5].
1. What the headline numbers mean: detainees versus arrests
Public statements and media reporting sometimes conflate “arrests” and “people in detention,” which changes interpretation. ICE’s biweekly detention snapshots and TRAC’s compilation show that 71% to 73.6% of people held in ICE detention in 2025 had no criminal conviction [1] [2]. By contrast, DHS press releases emphasize that a large share of ICE arrests involve people “charged or convicted” of crimes — a DHS claim that 70% of ICE arrests fell into that category appears in a July release [3]. These are different metrics: one is the criminality status of those currently detained; the other is DHS’s characterization of arrest categories. The data sources use different definitions and timeframes, so direct comparison without careful unpacking is misleading [1] [3] [2].
2. Independent analyses and watchdogs: the majority without convictions
Multiple independent organizations and news outlets using ICE’s own datasets conclude that people with no criminal conviction form the largest share of those detained. Migration Policy’s analysis reported that as of September 2025, 71% of ICE detainees had no criminal conviction [1]. TRAC’s November 2025 tally put the share at about 73.6% of detainees with no criminal conviction [2]. The Guardian’s reporting using ICE data similarly found more detainees with no criminal history than with convictions in a recent period [4]. The Brennan Center and Cato analyses likewise highlight that roughly two-thirds or more of those detained lacked criminal convictions in the cited windows [5] [6].
3. DHS and ICE messaging: “worst of the worst” versus selected statistics
DHS and ICE deploy selective statistics and narrative framing to emphasize public-safety rationales. DHS press releases assert that the agency is arresting “the worst of the worst” and have claimed that 70–75% of arrests were of accused or convicted criminals in specific periods [3] [7]. Those statements rely on ICE categorizations that can include people “charged” or “convicted,” and they typically describe enforcement priorities rather than the composition of the detained population at a point in time [3] [7]. Independent trackers caution that ICE’s published arrest counts may undercount some categories and that “collateral” arrests—when multiple people are taken into custody during an operation—inflate detainee numbers who have no criminal records [4] [8].
4. How definitions and timing shift the answer
Definitions matter: “arrested by ICE” can mean an interior administrative arrest, a transfer from CBP, or a courthouse arrest; “criminal conviction” can mean any past conviction (including minor traffic offenses) and “charged” is distinct from “convicted.” TRAC and Migration Policy use ICE’s detention roster snapshots to compute shares of detainees with convictions; DHS statements typically report arrest figures with broader categories [2] [1] [3]. Timeframes differ too: analyses covering the first 50 days of an administration produced different shares than biweekly detention snapshots taken later in the year [7] [1].
5. Competing interpretations and political context
Advocates and watchdogs argue the high share of people without convictions shows policy turned toward mass immigration enforcement and detaining people for civil immigration violations; they note many convicted detainees had only minor offenses (traffic, DUI, low-level drug or non-violent charges) [1] [2] [5]. DHS and ICE present enforcement successes against violent criminals to justify operations and to stress public-safety benefits [3] [9]. Both sides use ICE-produced data; the disagreement is over which subset of the data to emphasize and how to define “criminal.”
6. What reporting does not (yet) settle
Available sources do not mention a single, consistently applied national statistic that reconciles DHS/ICE arrest claims with detainee composition across the same timeframe and definition. In other words, there is no single, undisputed percentage that answers “what percentage of arrests by ICE are criminals” without specifying whether you mean arrests, current detainees, charged versus convicted, timeframe, or offense seriousness — and the available reporting shows these distinctions change the answer [1] [3] [2].
Conclusion: If you ask about the criminality of people in ICE detention during 2025, multiple independent counts show roughly 71–74% had no criminal conviction [1] [2]. If you ask about the share of ICE “arrests” that agency press releases characterize as involving people “charged or convicted,” DHS has claimed around 70–75% in its public statements [3] [7]. The apparent contradiction stems from differing definitions, timeframes and political framing; readers should compare the underlying tables used by ICE and independent analysts before accepting either headline.