Which U.S. regions or cities saw the largest increase or decrease in ICE arrests in 2025?
Executive summary
ICE arrests surged nationally in 2025, with advocacy analyses showing averages exceeding 1,000 arrests per day during peak periods and a marked shift toward community “at-large” arrests outside jails [1] [2]. The largest increases were concentrated in states that fully collaborated with federal enforcement and in specific regions—parts of the Deep South, the Rocky Mountain West, New Jersey, Maryland, Colorado, and some Southern California districts—while the Pacific Coast and greater New York area (notably northern California) registered relatively low arrest rates or smaller increases [1] [3] [4] [5].
1. High-collaboration states and the Deep South/Rocky Mountain clusters drove the biggest state-level rises
Analysts using ICE-provided datasets found that the sharpest statewide increases in arrests occurred where state and local governments fully cooperated with the administration’s enforcement priorities, creating geographic clusters of high arrest rates in the Deep South and parts of the Rocky Mountain region [1] [3]. The Deportation Data Project and UCLA’s state-variation work both show that state policy choices—not simply population—explain much of the variation in arrest volumes and rates, and that those cooperative states produced a disproportionate share of the nationwide surge [6] [3].
2. New Jersey, Maryland and certain Mid-Atlantic localities experienced outsized jumps
New Jersey emerged as a notable case where local collaboration with ICE produced sustained high arrest counts compared with states with stronger limits on cooperation like Illinois, Pennsylvania, or New York, and Maryland saw arrests nearly triple year-over-year through mid‑October 2025 according to the Deportation Data Project and local reporting [1] [7]. Local political fights over sheriffs and gubernatorial races in these states underscored that enforcement levels were responsive to local officials’ willingness to partner with ICE [1].
3. Urban and regional exceptions: Northern California stayed comparatively low while San Diego and others spiked
Northern California—centered on the San Francisco area of responsibility—maintained the nation’s lowest ICE arrest rate (about 217 per 100,000 noncitizens through mid‑October), well below the national average of roughly 1,000 per 100,000, even as national enforcement climbed [4]. At the same time, San Diego’s area of responsibility flipped from one of the lowest rates in 2024 to one of the highest in 2025, illustrating how enforcement intensity varies sharply within states and metropolitan regions [4].
4. Colorado and certain mountain/frontier counties saw notable increases tied to publicized operations
Colorado recorded over 3,000 ICE arrests through mid‑October 2025, with spikes tied to high‑profile operations in resort and Front Range areas and a shift in the detention source mix—many arrests came from community operations rather than transfers from the border or local jails [5]. Local data gaps, such as inconsistent facility coding, mean official counts likely understate specific operation sizes in places like Summit County [5].
5. Tactics changed in 2025; that reshaped which cities saw the biggest increases
National reporting and data analyses agree that ICE pivoted from jail-based arrests to at‑large community arrests, driven in part by pressure to increase daily arrest totals—this tactical shift amplified increases in metropolitan neighborhoods and suburban areas where ICE conducted community raids [2] [1]. That change meant cities and counties that previously saw low jail-derived ICE activity could nonetheless record large year‑over‑year arrest increases once community operations rose [2] [1].
6. Data limitations, alternate explanations and lingering questions
All of these conclusions rest on datasets compiled from ICE releases and FOIA-driven repositories that currently cover enforcement through mid‑October 2025 or later analytic snapshots; gaps, inconsistent facility labeling, and ICE’s limited public facility reporting complicate precise city-by-city tallies and trend attribution [6] [8] [5]. Federal OHSS monthly tables provide a complementary official feed but are structured around operational event counts rather than always mapping cleanly to local geographies [9]. Independent trackers (Deportation Data Project, TRAC) and reporting (Washington Post, local outlets) converge on the broad pattern—collaboration and tactical change produced the biggest increases—but local undercounting and political incentives in reporting mean some city-level extremes may be under- or over‑stated [6] [10] [2].