What proportion of ICE detainees are legal permanent residents or visa holders?

1. What the numbers actually show: detention vs. criminal history

Recent datasets and reporting emphasize criminal‑record status, not immigration status. TRAC’s snapshot shows 73.6% of the 65,135 people in ICE detention lacked a criminal conviction as of Nov. 16, 2025 ^[1]. The New York Times’ analysis of government arrest records similarly finds the share of people detained by ICE with any past conviction dropped to about 28% by mid‑October 2025 — implying roughly 72% had no convictions in that period ^[2]. Those numbers are the clearest, repeatedly reported metrics in the available sources; they do not directly answer the user’s question about proportions who are LPRs or visa holders ^{[1] [2]}.

2. What sources explicitly say about legal-status subgroups — and what they don’t

Opinion pieces and case reporting describe people “in the United States legally” being detained, and the New York Times opinion cites Migration Policy Institute data that 71% of those held at end of September lacked criminal convictions, noting many detained had lawful immigration histories ^[3]. But none of the provided sources supplies a clear, quantified breakdown of ICE detainees by lawful permanent resident status or by non‑immigrant visa holder category across the detained population ^{[3] [2] [1]}. Therefore, a precise proportion of detainees who are LPRs or visa holders is not found in current reporting.

3. Why that gap in reporting matters — different policy and legal implications

The distinction between someone with no criminal conviction and someone who is lawfully present is material: detention of LPRs or non‑immigrant visa holders raises different legal questions than detention of people without lawful status. The reporting emphasizes ICE’s increased arrests and the falling share of detainees with criminal records under the current administration, but it does not systematically map those people onto immigration categories such as LPR, temporary visa, asylum seeker, or unauthorized entrant ^{[2] [1]}. That lack of cross‑tabulated public data constrains civic oversight and legal analysis.

4. What reporters and advocates document anecdotally

Journalistic and opinion accounts highlight individual cases of people who were legally in the U.S. but still detained — for example, the New York Times opinion piece describing three people who “came to the United States legally” and were detained ^[3]. These narratives underscore policy and enforcement decisions that can sweep up people with lawful status, but anecdotes do not substitute for population‑level statistics ^[3].

5. Government framing and its limits

ICE and DHS material in the available set frame detention as enforcement against removable noncitizens and emphasize legal standards and custody management, but those official sources in this collection focus on process, detention standards, and enforcement aims rather than publishing a public breakdown of detainees by visa or LPR status for the dates in question ^{[4] [5] [6]}. DHS also issued rebuttals to some reporting about citizens being deported, underscoring agency sensitivity to narrative and the political stakes around detention statistics ^[7]. Those rebuttals do not provide the missing proportional breakdown on LPRs or visa holders.

6. Bottom line and where to look next

Bottom line: available sources document that roughly 70–74% of ICE detainees in the cited snapshots had no criminal conviction ^{[1] [2]}, and journalists report cases of legally present people being detained ^[3]. However, the exact proportion of detainees who are lawful permanent residents or nonimmigrant visa holders is not given in these sources — that specific statistic is not found in current reporting. For a definitive answer, seek ICE or DHS datasets that cross‑tabulate custody records by immigration status (LPR, visa class, asylum applicant, unauthorized) or request custom data from TRAC, Migration Policy Institute, or ICE’s statistics pages ^{[8] [4] [1]}.

Limitations: this analysis uses only the provided sources; available reporting emphasizes criminal‑record status and individual legal‑status anecdotes but does not publish the precise LPR/visa-holder share of the detained population ^{[1] [2] [3]}.