State‑by‑state: which jurisdictions saw the largest increases in ICE arrests of non‑convicted migrants in 2025?
Executive summary
The most reliable reporting and compiled ICE datasets indicate that the largest increases in arrests of non‑convicted migrants in 2025 clustered in a handful of cooperating jurisdictions—most notably Texas, Florida and California—with sharp local spikes in places like the Dallas field office; other states with increased interior arrests include New Jersey and parts of the Upper Midwest where new detention capacity was deployed [1] [2] [3] [4] [5]. Available sources show national surges in non‑convicted detentions but stop short of publishing a single, complete state‑by‑state percentage ranking in the materials provided here, so the granular ordering below assembles patterns reported by multiple outlets rather than a definitive numeric league table [5] [6].
1. Texas — epicenter by volume and local office surges
Multiple analyses identify Texas as one of the largest single contributors to the 2025 spike in ICE arrests and in new interior detentions, driven by cooperating county jails and large field‑office operations [1] [2]; local reporting on the Dallas office documents a dramatic rise in people without convictions booked in 2025 — jumping to “over 2,400” from under 500 the prior year — a clear signal Texas jurisdictions saw among the largest local increases in arrests of non‑convicted migrants [3].
2. Florida — cooperating state with rising jail‑based arrests
Florida is repeatedly named alongside Texas and California as accounting for a large share of arrests during the first months of Trump’s second term, with reporting and dataset mapping pointing to elevated arrest volumes tied to jail cooperation in that state [1] [2]. National analyses that highlight the role of cooperating states in concentrating ICE activity single out Florida as a primary locus of the skyward trend in interior arrests and detentions [5] [2].
3. California — high absolute numbers, mixed local patterns
California ranks among the states with the highest numbers of ICE arrests by mid‑2025, contributing to the triad (Texas, Florida, California) responsible for over 41% of early‑term arrests per the Deportation Data Project mapping, and thus saw significant absolute increases in arrests of migrants without convictions even where local policy resistance varies across counties [1] [5].
4. New Jersey and pockets with elevated shares of pending (but unconvicted) charges
State‑level breakdowns show different compositions of arrested populations; for example, New Jersey registered a larger share of ICE arrestees who had pending charges but no prior convictions compared with New York, indicating jurisdictional differences in the rise of non‑convicted arrestees and that New Jersey experienced notable increases in non‑convicted categories [5].
5. Upper Midwest / Minnesota region — detention logistics driving local increases
Reporting on ICE’s expansion of detention networks and use of new or repurposed facilities in the Upper Midwest — including plans around Minnesota’s Prairie Correctional Facility — indicates increases in arrests and transfers in that multi‑state region as capacity was stood up, suggesting localized surges tied to detention logistics rather than only to border proximity [4].
Context, competing narratives and data limits
Researchers and press analyses converge on a clear broad pattern: more arrests overall in 2025 and a rising share of those arrested had no criminal convictions or only pending charges, with national counts of non‑convicted detainees jumping sharply (for example, detained non‑convicted counts rose from 3,165 in February 2025 to over 25,000 by January in one analysis) [6] [7]. Yet federal messaging from DHS and ICE emphasizes a continuing focus on criminal aliens (DHS claims 70% in one release), a framing that conflicts with independent dataset trends and reflects institutional and political agendas that shape how statistics are presented [8]. Importantly, the sources available to this briefing document state‑level concentrations and local hot spots but do not provide a single, fully disaggregated state‑by‑state percentage increase table within the provided materials; the above assessment synthesizes repeated state mentions and local office reporting across the dataset rather than offering a mathematically complete ranking absent the raw state‑by‑state tables [5] [1] [3].