How does average ICE detention length differ by state and facility in 2025?
Executive summary
Average ICE detention lengths in 2025 vary widely — from a few days in many county jails to multiple weeks or months in large, dedicated immigration centers and private prisons — and the pattern differs by both facility type and state; however, facility-level and state comparisons are complicated by gaps and inconsistencies in ICE’s public reporting and by independent analyses that produce different national aggregates [1] [2] [3] [4].
1. How variation shows up: short stays, long stays, and everything between
Facility-level data assembled by researchers and nonprofits make clear that ICE’s “average length of stay” is not a single national constant but a distribution: many county lock-ups show average stays measured in days, while specialized immigrant detention centers and private prisons often report averages measured in weeks or months [1] [5]. The Vera Institute’s interactive dataset documents ICE use of 1,464 different facilities across 2008–2025 and shows that only a subset remain active in any given period, which produces stark facility-to-facility differences in average stays depending on purpose and population served [2].
2. Facility type drives most of the difference — county jails vs. dedicated centers
County jails and temporary lock-ups used for initial processing tend to register very short average lengths of stay (TRAC found many county facilities with “a few days” average length), while ICE-contracted dedicated detention centers and private prisons hold people for much longer periods because they house people awaiting removal proceedings or removal flights that can take weeks to months [1] [5]. ICE’s own detention-management framing emphasizes that custody decisions and the agency’s use of limited detention resources shape who is held and for how long, which reinforces why facility function correlates strongly with length of stay [6] [4].
3. State-level patterns: hubs, high-throughput sites, and regional differences
States that host major processing hubs, large private facilities, or border transit points tend to show higher average lengths and larger detained populations — examples in reporting include Texas and Louisiana’s heavy use of detention infrastructure and high throughput at certain airport or border-area facilities — while many interior states primarily show short-term jail holds or transfers out [7] [8] [9]. Vera’s facility map and the Deportation Data Project’s repository enable state-by-state and facility-by-facility exploration, but the headline is regional concentration: a small number of high-capacity facilities in a few states account for outsized volumes and longer stays [2] [10].
4. What 2025’s headline numbers say — and how to read them
Different outlets and datasets report different national averages: one independent aggregator reported an average length of stay moving from roughly 52 days to about 46 days in 2025, while other analyses emphasize record-high detained populations [11] [7]. ICE publishes dashboards with detention and custody data through the ERO statistics pages, but those dashboards are built on ICE’s chosen subsets and methodologies [4]. Independent watchdogs and the GAO have documented that ICE’s public reporting understates total detentions because of methodology choices (for example, excluding certain temporary bookings), which complicates deriving clean state or facility averages from official releases alone [3].
5. Data caveats that should shape any comparison
Comparing average lengths across states or facilities in 2025 requires care: facility definitions change, ICE contracts open and close (Vera notes only about one-third of facilities were active mid‑October 2025), datasets have missing fields, and federal reporting choices lead to undercounts [2] [3]. Advocacy organizations likewise highlight that long stays correlate with reported abuses and medical neglect in some facilities, which adds an ethical dimension to any statistical comparison but does not by itself yield a standardized metric for “average length” across contexts [12] [5].
6. Bottom line: disparate experiences, one patchy dataset
The empirical pattern in 2025 is firm: average detention lengths differ substantially by facility type and by state, with short averages in many county jails and substantially longer averages in specialized centers and private prisons; nonetheless, precise state-to-state rankings or a single nationally representative average are constrained by ICE’s reporting choices and by gaps identified by external auditors and researchers, so analysts must rely on facility-level data compilations (Vera, Deportation Data Project, TRAC) and treat official aggregates with documented caution [2] [10] [1] [3].