ICE Detention and Deportation by the Numbers
ICE published updated detention data last week. Get a full explanation of the data below (great for new readers), see my revised approach to data viz, and follow along as I include even more data.
ICE published new detention data in early April for the first time since February, ending a 56-day gap during which the agency didn’t publish any new data. The new release covers data through April 4, 2026, and includes the first look at deportation and detention figures covering most of the winter. Below is a breakdown of some of the key numbers, starting with deportations and working down to the specifics of who is being booked into detention and by which agency.
You might be saying, “wait, you already wrote about this?”—and you’d be right. But I’m making some adjustments to how I process and visualize the data, and I’m expanding to include more data to make the most of what ICE is publishing. You can help support my commitment to constant improvement by sharing this with others and learning about your options for becoming a paid subscriber.
ICE Removals
The chart below shows average ICE removals per day from early FY2024 through April 4, 2026. Each bar represents one reporting interval (the period between two consecutive biweekly data releases) and its height is the average number of people deported per day during that interval. The bars are not all the same width because the time between releases varies, sometimes spanning two weeks, sometimes longer when ICE delays or skips a release. The large gap in early 2026 reflects the period when ICE stopped publishing data entirely.
The most recent interval (56 days ending April 4, 2026) recorded 72,044 total removals, an average of 1,286 per day. The peak so far in the dataset was 1,456 per day in late January 2026. The trend over the past several months has been roughly flat in the 1,250–1,400 range, which is where it has been since late summer 2025.
Through the first six months of of FY2026 (October 1, 2025 through April 4, 2026), ICE has carried out 234,236 removals. At the same point in FY2025 (through April 5, 2025), the total was 134,500. At the same point in FY2024 (through April 6, 2024), it was 133,803. In other words, the current administration has deported roughly 74% more people in the same stretch of the fiscal year than either of the two prior years. Looking back, FY2025 ended with 319,980 total removals; FY2024 ended with 248,739. At the current daily pace of 1,266 per day, FY2026 is on track to exceed 460,000 removals for the full year, nearly 45% more than FY2025 and 85% more than FY2024.
There are three visible gaps in the chart, and they have different explanations. The two shorter gaps—one running from September 7 to September 30, and another from September 20 to September 30—are end-of-fiscal-year gaps that recur annually. ICE’s fiscal year runs from October 1 through September 30, and the agency’s YTD removal totals reset to zero at the start of each new fiscal year. The best explanation to my knowledge (please correct me) is that ICE gets busy with year-end reporting obligations in late September and early October, and biweekly releases are routinely delayed or skipped during that window. This means there is no data for the roughly six weeks between the last September release of one fiscal year and the first release of the next, although the data does backfill to October 1. These are expected structural gaps in the data, not anomalies.
The third long bar—56 days running from February 7 to April 4, 2026—has a different cause. ICE stopped publishing its biweekly statistics entirely during this period, apparently tied to disruptions during the Congressional standoff over DHS funding. The agency is legally required under the Homeland Security Act to publish this data biweekly, and its failure to do so prompted public concern from researchers and advocates. I flagged this publicly when the data stopped getting published. The practical effect is that we have a single 56-day data point covering most of the winter rather than the usual four or five two-week snapshots: we know the total (72,044 removals over that period) but can’t see any week-by-week variation within it.
The Trump administration’s stated goal is 1 million removals per year, language that appears verbatim in ICE’s congressional budget justification for FY2026, making it an official policy target and not just a campaign talking point. FY2026 is on pace for roughly 460,000 ICE removals at the current daily rate, less than half the official goal but since the administration is fudging their deportation numbers, it’s hard to tell if this trivial factual matter will stop the administration from making apples-to-oranges-style claims.
One definitional note worth keeping in mind: according to ICE’s own footnotes in the spreadsheet, the removal figures reported here include returns and expulsions, with returns defined as voluntary returns, voluntary departures, and withdrawals under docket control. The figures also include people processed for expedited removal or voluntary return who were turned over to ICE’s Enforcement and Removal Operations for detention or removal, and, starting in March 2025, Title 50 expulsions (i.e., Alien Enemies Act expulsions). They do not include people processed for expedited removal who were not detained by ICE and were handled primarily by Border Patrol. In short, the figures are broader than a strict count of formal deportation orders but narrower than the aggregate “deportations” figure the administration has cited publicly. DHS recently claimed a cumulative total of 622,000 deportations, a number I’ve written about separately because it bundles together legally and procedurally distinct actions in ways that make direct comparison difficult. The figures in the chart above come from ICE’s own published spreadsheets and use ICE’s own definitions.
ICE’s Detained Population
The chart below shows the total number of people held in ICE detention on the date of each biweekly release: a point-in-time snapshot rather than a cumulative count. Each data point reflects a single day’s detention population across all ICE facilities nationwide, not including people held in short-term CBP facilities, field offices, or courtrooms (an historically minor issue that has become a major transparency problem).
The detained population peaked at 70,766 on January 24, 2026, the highest number ever recorded in ICE’s publicly available data, and the first time the agency has held more than 70,000 people. I wrote about that record when it was published. By the April 4 release, the population had fallen to 60,311, a decline of 10,455 people, or about 15%, over roughly ten weeks. That decline is real, but it’s worth keeping in perspective: 60,311 is still higher than any detention population recorded before this administration and some detention centers have actually seen growth. Subscribe to get a forthcoming report coauthored with Adam Sawyer.
The bulk of the decline came from a 21% drop in detentions of people with no criminal history, a group ICE classifies as “Other Immigration Violators.” That drop followed the widely reported controversy over ICE’s enforcement operation in Minneapolis, where agents shot and killed two civilians. The bipartisan backlash appears to have produced a measurable, if temporary, shift in agency behavior. The administration has announced no policy change, and there’s no particular reason to expect the decline to continue.
Camp Deaths
Since the start of the second Trump administration in January 2025, 47 people have died in ICE custody, based on data compiled from ICE’s own press releases. Sixteen of those deaths have occurred in 2026 alone, at a rate of one every 6.3 days. The most recent was Alejandro Cabrera Clemente, a 49-year-old Mexican national who died on April 11 at Winn Correctional Center in Winnfield, Louisiana, a privately operated facility with a documented history of inadequate medical care. Before him, Tuan Van Bui died on April 1 at Miami Correctional Center in Indiana. At the current pace, 2026 is on track to exceed 60 deaths in ICE custody for the year.
Are ICE detention centers “camps”? Watch my conversation with John Washington and tell us what you think.
The agency’s public response to these deaths has been to minimize them. ICE press releases describing custodial deaths have introduced language that appears nowhere in official policy: a recent release stated that ICE detention provides “the best healthcare than many aliens have received in their entire lives”—a claim that is both ungrammatical and contradicted by the deaths themselves, as well as by accounts from detainees, attorneys, and family members. A Haitian asylum seeker died of a dental infection at a facility making that claim. Rather than investigate the systemic conditions producing these deaths, the administration has used press releases to criminalize the deceased and deflect responsibility.
To learn more about ICE’s custodial deaths, read #DetentionKills by Andrew Free.
Who Is Being Detained: Criminal History
This chart shows the detained population broken down by criminal history, but only among people arrested by ICE (interior enforcement arrests, not CBP border arrests). ICE classifies people into three categories: convicted criminals, people with pending criminal charges, and “Other Immigration Violators,” meaning people whose only violation on record is an immigration offense. Criminal history is one of the most important dimensions of this data because the Trump administration has consistently justified its enforcement operations by claiming to target dangerous criminals and public safety threats. If that justification holds, you’d expect the convicted criminal category to dominate the detained population. It doesn’t.
People with no criminal record beyond an immigration violation have consistently made up the largest share of ICE’s detained population and account for most of the growth and decline over the past year and a half. The Trump administration has argued publicly and repeatedly that its enforcement priorities are focused on people who pose a serious public safety threat. The data presents a different picture. A Cato Institute analysis of ICE detention data covering the first weeks of FY2026 found that 73% of people in ICE custody had no criminal conviction at all, and only 5% had a violent conviction. The criminality breakdown in this chart is consistent with that finding: people with convictions remain the smallest and slowest-growing segment of the population.
Two recent enforcement incidents illustrate how this plays out on the ground. ICE detained two Venezuelan emergency room physicians in south Texas within days of each other (the New York Times reported on the second case on April 11), with no public explanation of what public safety risk either individual posed. Separately, immigration judges were dismissed the same week after issuing rulings that blocked the deportation of pro-Palestinian students. The executive branch’s response to judicial review that slows enforcement has been to remove the judges.
Book-Ins by Arresting Agency
These two charts show the same data in two formats (a grouped bar chart and a line chart) of average daily book-ins by arresting agency. A “book-in” is a person formally entered into ICE detention. This is a flow measure: it counts how many people enter the system, not how many are held at any given moment. I’ve constructed a monthly time series going back to January 2024, using the most complete available data for each fiscal year, and presented it as a daily average to deal with the issue of fractional months. Months where the data is still partial (meaning ICE’s reporting period hadn’t closed out the full month yet or the end of the fiscal year) are flagged with hatched bars or open markers.
In the months before the inauguration, ICE was booking in roughly 253–379 people per day. By February 2025 (the first full month of the new administration) that number had jumped to 621 per day, and it continued climbing through the summer, peaking at 1,053 per day in June 2025. Since then it has settled into a range between 930 and 1,220 per day. March 2026, the most recent complete month, came in at 955 ICE arrests per day.
CBP book-ins tell the opposite story. During 2024, CBP was contributing between 122 and 485 people per day to ICE detention, a substantial share of the total intake that reflected the high volume of border encounters at the time. Since January 2025, those numbers have fallen sharply. By March 2026, CBP was contributing just 94 people per day, down from a high of 485 in October 2023. The current total daily intake (ICE plus CBP combined) is around 1,049 per day, and ICE now accounts for about 91% of it.
The April 2026 bars are hatched and the April markers are open because they cover only the first four days of the month (April 1–4). The per-day rates for April (930 for ICE and 108 for CBP) are based on a very small sample and shouldn’t be read as a trend signal yet.
Follow for Updates, Circulate Graphs with Attribution
All of these figures come from ICE’s biweekly detention statistics, which the agency publishes under a legal obligation established by the Homeland Security Act. The data is current as of April 4, 2026. I’ll update these charts each time ICE releases new data. For facility-level detail, including population trends at individual detention centers, visit DetentionReports.com, a data project I run with Adam Sawyer at Relevant Research.
Any of the charts or data from this post are free to use with attribution, no need to reach out first. If you reach out with questions or support related to immigration data analysis presented here, consider becoming a paid subscriber.
i would really like to attend your upcoming zoom about data, but my schedule won't work with that time. Will it be recorded and available later, I hope?
Government shutdowns in October