How many illegal immigrants use food stamps?

1. Why the question keeps circulating: political claims vs. data

Political leaders and social-media posts have pushed the claim that immigrants — especially “illegal” or undocumented immigrants — make up the majority of SNAP recipients; journalists and fact‑checkers say those claims are misleading or false because federal data show the vast majority of SNAP beneficiaries are U.S. citizens, and the Center for Immigration Studies’ 59% statistic was misinterpreted from a study about households headed by immigrants using any major program, not the share of SNAP recipients who are undocumented ^{[6] [2] [3]}.

2. What the official rules say about undocumented immigrants and SNAP

Federal SNAP rules historically have excluded undocumented immigrants from eligibility; certain categories of non‑citizens (refugees, asylees, some lawful permanent residents after waiting periods, etc.) can qualify, but undocumented people are not eligible for federal SNAP benefits ^{[1] [7]}. The Food and Nutrition Service and USDA pages under the One Big Beautiful Bill (OBBB) of 2025 note changes to non‑citizen eligibility and direct states to verify immigration status via SAVE, reflecting recent regulatory shifts that states are implementing ^{[4] [7]}.

3. How researchers and fact‑checkers measure participation

Analysts use different data sources and units — people, households, or households “headed by” immigrants — and that matters. The Center for Immigration Studies used the 2022 Survey of Income and Program Participation to report that 59% of households headed by immigrants used “at least one major program,” which was widely mischaracterized on social media as “59% of food stamp recipients aren’t American” ^{[3] [2]}. PolitiFact and WRAL both highlight a different metric: about 17% of households headed by undocumented immigrants participated in SNAP, and USDA 2023 figures show almost 90% of SNAP recipients are U.S.-born citizens and 98% of SNAP-participating households include U.S. citizens ^{[3] [2]}.

4. Recent administrative actions that affect counts and perception

In 2025 the USDA under Secretary Rollins ordered enhanced verification practices and directed states to use SAVE to confirm identity and immigration status — an explicitly political and administrative stance meant to limit any improper benefit receipt and to signal stewardship of taxpayer dollars ^[8]. Simultaneously, the OBBB enacted July 4, 2025 altered non‑citizen SNAP rules and required states to apply new criteria immediately for new applicants and at recertification for existing cases; FNS issued memos and a 120‑day variance exclusion tied to November 1, 2025 implementation timing ^{[4] [7]}.

5. The limits of available reporting and what we don’t know

Available sources do not provide a single, current national tally specifically enumerating the number of undocumented immigrants receiving SNAP as of November 2025; instead, reporting relies on surveys, administrative categories, and household definitions that produce different percentages ^{[3] [2]}. Where one analysis counts “households headed by undocumented immigrants that use any program,” others measure shares of SNAP recipients by citizenship status — these are different comparisons and lead to different headlines ^{[3] [2]}.

6. What to take away: headline facts and caveats

Headline facts supported by reporting: undocumented immigrants are generally not eligible for SNAP ^[1]; USDA data show most SNAP recipients are U.S. citizens (about 90% U.S.-born and 98% of participating households include U.S. citizens in cited USDA data) ^[2]. Caveats: smaller shares of immigrant‑headed households do participate in SNAP, often because households are mixed‑status (U.S.-born children with noncitizen adults), and policy changes in 2025 are shifting verification and eligibility practices, which may change future counts ^{[3] [4] [7]}.

If you want, I can compile the specific numerical claims and the original source lines (Center for Immigration Studies, the USDA reports, and the fact‑checks) into a side‑by‑side comparison so you can see exactly how differing definitions produce very different percentages ^{[3] [2] [6]}.