How many people did ICE deport each year from 2017 to 2025, broken down by removal vs. returns?

Checked on December 2, 2025
Disclaimer: Factually can make mistakes. Please verify important information or breaking news. Learn more.

Executive summary

Available public datasets and news analyses show that ICE and DHS report removals (formal deportations) and returns (voluntary or expedited departures) in different ways and that counting methods changed in 2023–25, complicating year‑to‑year totals (ICE notes data fluctuate until fiscal-year “lock” and released files differ) [1] [2]. Independent trackers and reporters — including Migration Policy Institute and the Deportation Data Project — document that ICE removals from the U.S. interior fell after 2016 and that DHS/ICE reporting through 2024–25 has included both agency‑initiated removals and larger “self‑deportation” or voluntary return estimates, producing widely varying headline numbers [3] [4] [5].

1. Why a simple “2017–2025” yearly table is not available

ICE and DHS publish removals and returns in multiple datasets and formats; ICE cautions its numbers “fluctuate until ‘locked’” at fiscal‑year close and the agency changed its removals‑counting methodology in mid‑2023, applied retroactively to May 12, 2023 — creating discontinuities researchers flag as important [1] [2]. The Deportation Data Project and other analysts advise relying on specific ICE FOIA files (late‑July 2025 release) because earlier files for FY2025 were incomplete or inconsistent [4] [2].

2. What “removals” vs. “returns” mean in official data

ICE/ERO distinguishes “removals” (formal deportations following a removal order or ICE transport) from “returns” or “voluntary returns” and from “expedited removal,” and some DHS datasets mix or exclude subsets depending on how the person entered ICE custody [1] [2]. The OHSS MONTHLY TABLES (DHS statistical system of record) provide Removals and Returns breakdowns by citizenship and criminality, but those tables are compiled from operational reports and are updated monthly — not a single static yearly release [6].

3. The long‑term trend analysts emphasize

Pre‑2017, annual interior removals averaged higher: MPI reports interior removals averaged about 155,000 per year in FY2009–16, falling to roughly 81,000 per year in FY2017–20 — demonstrating a structural decline in ICE interior removals after 2016 [3]. That provides context: year‑to‑year shifts after 2017 are changes on top of a multi‑year decline [3].

4. Why 2024–2025 headline numbers diverge widely

Multiple outlets and trackers show a surge in reported enforcement activity in 2025 but differ on what counts as an ICE deportation versus a DHS estimate of departures. The Trump administration and DHS circulated large figures including “voluntary self‑deportations” (1.6 million cited by some officials and linked to an external estimate from the Center for Immigration Studies), while independent reporting and FOIA datasets focus on ICE removals and ICE’s own FOIA‑provided datasets show far lower counts for removals alone; scholars and outlets warn these different concepts should not be conflated [5] [7] [8]. The Deportation Data Project and news organizations (The Guardian, Newsweek) have published FOIA‑based ICE removals files through 2025 but note incomplete files and changing methodology [4] [9] [8] [2].

5. What public sources do provide (how to get a year‑by‑year breakdown)

If you need a year‑by‑year removals vs. returns table for 2017–2025, use two primary sources: ICE’s published Enforcement and Removal Operations statistics (removals/returns tables) and DHS/OHSS “Immigration Enforcement Monthly Tables” (Removals and Returns by Citizenship) which together allow fiscal‑year aggregation — but be aware ICE’s FY2025 datasets required corrections and researchers recommend specific FOIA files (late July 2025) for accuracy [1] [6] [2]. The Deportation Data Project hosts curated ICE FOIA releases and commentary useful for assembling consistent series [4].

6. Competing perspectives and hidden agendas to watch

Government press releases and administration briefings have used aggregated “departures” and outside estimates (e.g., CIS estimates) to frame large numbers that include voluntary departures; critics and independent researchers caution those figures inflate the footprint of ICE‑carried removals [5] [7]. Advocacy groups, trackers and ICE itself have different incentives: ICE seeks to report operational output; advocacy groups and journalists focus on methodology and human‑rights implications, while some think tanks supply population estimates that administrations may repurpose as enforcement metrics [4] [5].

7. How I’d proceed to produce the exact table you asked for

I can compile a year‑by‑year table for FY2017–FY2025 using ICE’s “Removals and Returns” files and DHS OHSS monthly tables, selecting the late‑July 2025 ICE removals file as recommended by the Deportation Data Project to minimize known gaps [1] [2] [6]. Available sources do not include a ready, validated 2017–2025 single table in your search results; tell me if you want me to assemble that table from the named files and I will extract and cite each yearly removals and returns number from those datasets [1] [6] [2].

Want to dive deeper?
What were ICE's total apprehensions and referrals each year from 2017 to 2025?
How do DHS/ICE definitions of removals versus returns differ and affect statistics?
Which countries received the most removals and returns from the U.S. annually since 2017?
How did policy changes under each presidential administration impact ICE removal numbers from 2017–2025?
Where can I find the primary DHS/ICE datasets and FOIA sources to verify annual removal vs return counts?