MEBRO

DISINFO DESK

Science & Health

No, Florida Didn't Just Find Your Halloween Candy 'Laced' With Arsenic — That Test Is From January, and No Federal Limit Says It's Unsafe

A viral post warned parents that kids could overdose on arsenic from Kit Kats and Snickers this Halloween, citing a Florida study that found arsenic in 60% of tested candies. The number is real — but the study is nine months old, the 'safety limit' is the state's own invention, and a toxicologist says Florida used the wrong test.

MISLEADING ≈

FILED NOV 3, 2026 · UPDATED NOV 3, 2026 · 3 SOURCES

The Claim

In late October 2026, researcher Nicolas Hulscher posted on X that children were about to collect "arsenic-laced candy" for Halloween, citing a claim that Florida had found arsenic in 60% of popular candy brands and that kids could eat only a handful of pieces of candies like Kit Kat or Snickers per year before exceeding an "arsenic safety limit" [1]. The post timed a months-old state report to the holiday, reviving numbers that had already begun recirculating online around Oct. 16, 2026 [1]. The 60% figure is accurate — but it describes a January test, not a fresh Halloween-season finding.

The Evidence

Florida's Department of Health announced the results on Jan. 26, 2026, under Gov. Ron DeSantis's "Healthy Florida First" initiative: an independent lab tested 46 candy products from 10 companies bought online and in Florida stores, and detected arsenic in 28 of them — about 61%, which rounds to the viral post's "60%" [1][2]. Reported levels ran roughly 180 to 570 parts per billion depending on the product, with items like Jolly Ranchers and Nerds varieties among the higher readings [2][3]. The state used EPA Method 6010D, a technique the whatthefact.app analysis and FOX 13 News both note is typically applied to soil and groundwater, not food [1][3]. University of South Florida toxicologist Dr. Marie Bourgeois told FOX 13 News, "If they're using something for soil, when they should be using something for food, then you're not getting the right answer" [3]. That method measures total arsenic without separating the far more toxic inorganic form from the much less harmful organic form — a distinction the American Cancer Society treats as central to risk, and one Florida's results don't make [1][3].

Historical FDA testing of candy has generally found arsenic at or below 15 parts per billion — far under Florida's reported hundreds — which is part of why outside toxicologists are questioning the state's numbers rather than accepting them at face value [3]. No recall was issued, and the FDA has no arsenic action level for candy at all [1][2].

Who's Spreading It

The "safety limit" driving the viral alarm isn't a federal standard — it's a benchmark Florida calculated itself from hypothetical annual consumption, and it hasn't been published with its underlying sample sizes or detection limits [1]. Florida's own surgeon general, Joseph Ladapo, has called the limits "highly conservative" [3], while the National Confectioners Association disputes them outright, saying they don't align with federal standards or peer-reviewed science and that the industry's products are safe [1][3]. The same state official who set the alarming limit has also acknowledged it runs hotter than federal guidance would require.

The Verdict

Taken claim by claim: Florida did find arsenic in 28 of 46 candies tested, so the "60%" figure holds up [1]. But the framing that millions of children are about to be served "arsenic-laced" candy this Halloween is misleading — the test is nine months old, was not timed to Halloween, triggered no recall, and relies on a self-created "safety limit" rather than any FDA or EPA food-safety standard, built on a measurement method outside toxicologists say is designed for soil, not candy [1][2][3]. The number is real; the Halloween-panic framing built around it is not.

MEBRO · DISINFO DESK · mebro.app

Investigative report — not a user-submitted fact-check.

AI-built, source-verified. Every claim here was checked against the sources cited above before publishing — but don't just trust us: follow any citation to its source and confirm it yourself. That's the whole point.