A preliminary DHS data match found roughly 256,000 potential noncitizen registrants in California, New Jersey, Nevada and Pennsylvania, and the department wants state election officials to help verify the names before the 2026 midterms.
Homeland Security Secretary Markwayne Mullin sent letters Friday to election officials in four states laying out numbers that, if confirmed, would be among the largest noncitizen voter registration findings in recent memory. The DHS review found 190,832 potential noncitizen registrants in California, 35,152 in New Jersey, 15,903 in Nevada and 14,576 in Pennsylvania, according to the letters, first reported by Fox News and The Federalist. Mullin asked the four secretaries of state to contact DHS by July 24 so the department can begin sharing underlying records and start the verification process.
The method is straightforward: DHS matched state voter rolls against federal immigration databases looking for names that appear in both. President Trump cited the findings in a primetime address this week, framing them as evidence of vulnerabilities in the election system. That claim landed differently depending on who you asked.
DHS itself describes the review as preliminary. That word matters. Data-matching programs like this one have a well-documented flaw: naturalized citizens routinely get flagged as noncitizens because the immigration record used in the comparison predates their citizenship, and the database is never updated to reflect it. The federal government's own data-sharing agreements with states warn that officials must do independent due diligence before using match results to purge anyone from the rolls. A name showing up on this list is a lead, not a verdict.
Nevada's Secretary of State's office pushed back hard. A spokesperson called the 15,903 figure inflated and said the state's own citizenship verification checks, run through the DMV and other records, routinely catch and correct these mismatches long before a name would ever reach a ballot. California officials offered a similar response, noting that the state automatically registers eligible voters through the DMV under its motor voter law, a system built around license and identification records that already screen for citizenship status. Both states said they had not yet received the underlying data DHS says it will share, making it impossible for local officials to independently confirm any of the figures.
New Jersey and Pennsylvania have been more measured in their public statements, saying only that they would review the DHS letters and respond through appropriate channels. Pennsylvania in particular has faced scrutiny over its voter rolls before. A 2017 state audit found a small number of noncitizens had registered to vote over several years, mostly people who mistakenly checked a box during motor vehicle transactions rather than through any coordinated effort. That episode is often cited by both sides of the debate, either as proof the system catches errors or as proof the errors happen at all.
This is not the first time a federal data match has produced a large, attention-grabbing number that shrank considerably under review. Florida's 2012 effort to purge noncitizens from its voter rolls initially flagged more than 180,000 names, a list that was later whittled down to under 200 confirmed cases after county officials checked them individually. Texas ran into similar trouble in 2019, when its Secretary of State flagged roughly 95,000 voters as possible noncitizens using a comparison against Department of Public Safety records, only to retract the list within weeks after acknowledging thousands of naturalized citizens had been wrongly included. The state settled a lawsuit over the episode and paid the plaintiffs' legal fees.
Voting rights groups, including the ACLU and the Brennan Center for Justice, have already signaled they will be watching how DHS and the four states handle this latest match. Their concern is less about whether noncitizens are registered, which does happen in small numbers, and more about what happens next: whether officials use the raw DHS list to trigger removal notices or registration challenges before every name is individually verified. Election law in several states requires a multi-step notice-and-response process before anyone can be struck from the rolls, a safeguard designed specifically to prevent exactly this kind of unverified list from causing wrongful purges.
Mullin's letters do not direct the states to remove any names. They ask for cooperation in cross-checking DHS's data against state records, a process that in past cases has taken months. California, with by far the largest number of flagged registrants, has the heaviest lift ahead of it, and its response could set the tone for how seriously other states treat the findings.
With the July 24 deadline for state contact approaching and midterms still more than a year off, the real test will be in the verification numbers that come back once officials actually dig into the names. Given the track record of similar data-matching efforts in Florida and Texas, election experts caution that the final confirmed count of ineligible registrants is likely to be a small fraction of the 256,000 figure now making headlines. Whether that nuance survives the political conversation around it is another question entirely.
Also read: Trump administration revives green card rule denying immigrants who rely on welfare • House GOP unveils $95 billion reconciliation bill as hawks balk at no offsets • Rubio hosts 65 nations to fight far-left terror, more group bans coming
