How many criminal and dangerous immigrants has ICE arrested?
Executive summary
The answer depends on which metric is used: DHS and ICE tout that roughly 70% of arrests by ICE are of people with U.S. criminal charges or convictions (their “criminal illegal aliens” framing) while independent data-tracking and research organizations show that a large—and in many datasets the majority—of people in ICE custody or arrested by ICE have no U.S. criminal conviction, and only a small share have violent convictions [1] [2] [3] [4] [5]. Both claims are supported by different slices of ICE’s own reporting and third‑party analyses, so any definitive single number requires clarifying which “arrest” or “custody” population is being counted [3] [5].
1. DHS’s reported totals and the “70% criminal” claim
The Department of Homeland Security publicly asserts that about 70% of those arrested by ICE in this enforcement surge have criminal charges or convictions in the United States, and DHS press releases list high totals of arrests and removals—statements that frame the operation as focused on “the worst of the worst” [1] [6] [7]. DHS also declared totals such as “more than 670,000 removed in one year” in recent promotional materials, and its messaging highlights individual high‑profile criminal cases to illustrate that claim [6] [7]. Those DHS numbers refer to arrests and removals as counted by the agency and to categories ICE tracks in its Enforcement and Removal Operations datasets, which distinguish convictions, pending charges, and “other immigration violators” [3] [1].
2. Independent trackers: larger totals and a different breakdown
Independent trackers and watchdogs using ICE book‑in and detention data show much larger cumulative arrest and detention totals for the administration—The Guardian reported the administration had arrested more than 328,000 people and deported nearly 327,000 through part of FY2026—and organizations like TRAC report tens of thousands of ICE interior arrests in single months [8] [5]. Those datasets also reveal that a substantial share of people booked into ICE custody have no U.S. criminal convictions: TRAC and other analysts reported that large portions of the detained population had no convictions, and some analyses claim more than two‑thirds of arrests or detainees lacked a criminal conviction in U.S. courts [5] [9] [4].
3. Violent and “dangerous” convictions are a small share
Multiple independent analyses emphasize that the share of ICE detainees with violent convictions is low: the Cato Institute’s analysis of leaked/FOIA data showed roughly 5% of people detained had violent convictions, and other Cato calculations placed 65–67% of people taken by ICE as having no convictions and 93% lacking violent convictions in some releases [4] [10]. ICE’s own ERO statistics classify arrests by convictions, pending charges, and non‑criminal immigration violations—demonstrating that different categories produce very different headline percentages depending on the denominator chosen [3].
4. Why the numbers diverge: definitions, timeframes, and processing points
The disagreement is often methodological: DHS and ICE cite the share of “arrests” that involve someone with a criminal conviction or charge, which can include pending charges and U.S. immigration crimes; independent researchers often analyze the population actually held in ICE custody or booked into detention, where recent policy changes produced a surge of non‑criminal immigration arrests and a swelling detained population of people with no U.S. convictions [3] [9] [11]. Timeframe matters too—the agency’s cumulative arrest counts across operations can exceed book‑ins in any single month—and reliance on local jail referrals and variations in how “criminal” is defined (traffic offenses, immigration misdemeanors, pending charges) drives further divergence [12] [5].
5. Bottom line answer
There is no single uncontested headline number for “how many criminal and dangerous immigrants ICE has arrested” in public sources: DHS reports that about 70% of ICE arrests involve people with U.S. criminal charges or convictions [1] [2], but independent data analysts and watchdogs show that large absolute numbers of arrests and detentions involve people with no U.S. criminal convictions and that the share with violent convictions is small—on the order of single digits to the low teens depending on the dataset [4] [5] [9]. The clearest path to a precise answer requires specifying whether one means cumulative arrests by ICE, book‑ins to ICE detention, arrests with convictions versus pending charges, or arrests for violent offenses—because each yields a different, well‑documented figure in the sources above [3] [8] [5].