Is Titanium Dioxide Safe to Eat? Europe and the FDA Read the Same File and Disagree

You have almost certainly eaten titanium dioxide. It is the reason the white coating on a mint stays bright, why some frostings look opaque instead of translucent, why certain candies glow that particular flat white. You have probably also painted a wall with it, brushed it onto your skin as sunscreen, and squeezed it out of a toothpaste tube. It is the same compound in all of those places. That last part is where the calm stops being calming.

Here is the thing that makes titanium dioxide unusual among food additives: two of the most respected food-safety agencies on earth looked at the same pile of studies within about a year of each other and reached opposite conclusions. Europe pulled it from the food supply. Canada, the UK, and the United States kept it. Nobody is lying. That is what makes this a file worth opening.

The claim vs. the evidence

The claim you'll hear	What the evidence actually shows
"It's banned in Europe, so it must be poison."	The EU banned it in 2022 because its scientific panel could not rule out genotoxicity — not because harm was demonstrated. EFSA explicitly said evidence of general toxic effects "was not conclusive."
"The FDA says it's perfectly safe, so it's fine."	The FDA still permits it and says it did not identify genotoxicity concerns in the available data — but a petition to revoke that approval has been under review since 2023.
"It's a weird industrial chemical in food."	True and beside the point: it's a white pigment, the same TiO₂ used in paint. Whether that matters depends on particle size and dose, which is exactly what the agencies argued about.
"The nanoparticles are the smoking gun."	Up to half of E171's particles are nano-scale. EFSA treated that as a reason for caution. Health Canada argued food-grade exposure behaves differently than the lab studies suggest. Genuinely unresolved.

What it actually is

Titanium dioxide is a white pigment. Not a dye. A dye dissolves and tints; a pigment scatters light and makes things bright, white, and opaque. Chemically it is TiO₂, and it is one of the most-used white pigments in the world — roughly 4.6 million tonnes of pigmentary titanium dioxide are produced annually, the overwhelming majority of it going into paints, coatings, plastics, and paper. Food coloring, where it carries the European code E171, is a rounding error by comparison.

The "also found in sunscreen and paint" line gets used to alarm people, and it shouldn't, on its own. Salt is also used to de-ice roads. The question is never whether a substance appears somewhere industrial. The question is dose, form, and what happens once it's inside you. So let's go to the file.

How it got approved

The FDA permanently listed titanium dioxide as a color additive for food on January 27, 1966 (31 FR 1065), with the rule taking effect that October. The limit set then is the limit now: titanium dioxide may be used to color food generally, provided the quantity "does not exceed 1 percent by weight of the food" (21 CFR 73.575). It is also exempt from batch certification — the FDA decided that testing each batch before sale was "not necessary for the protection of the public health."

That logic made sense in 1966. The tools available to detect what a particle does at the nanometer scale, inside gut tissue, over a lifetime, mostly did not exist yet. What happened over the next fifty years did not stay still — the science of nanoparticles arrived, and it arrived pointing straight at this compound. (For how the broader U.S. additive-approval system lets decades-old clearances persist untouched, we filed a separate report on the FDA's GRAS pathway.)

The receipts

On May 6, 2021, the European Food Safety Authority published a formal opinion (EFSA Journal 2021;19(5):6585). Its expert panel applied EFSA's 2018 guidance on nanomaterials to E171 for the first time and found that up to 50% of its particles fall in the nano range — smaller than 100 nanometers. The conclusion was blunt: an Acceptable Daily Intake "cannot be established," and titanium dioxide "can no longer be considered safe when used as a food additive." Panel chair Prof. Maged Younes put it plainly, and the panel's reasoning rested on a specific and narrow point, stated by panel member Prof. Matthew Wright: "we could not rule out a concern for genotoxicity" — the possibility that the particles could damage DNA.

Read that carefully, because the brand promise here is that we don't inflate it. EFSA did not say titanium dioxide causes cancer. It said that, after reviewing the data, it could not exclude the possibility of genotoxicity, and that when you cannot exclude genotoxicity, you cannot set a safe daily dose. That is a precautionary standard, and EFSA was explicit that evidence of general toxic effects "was not conclusive."

Europe acted on it anyway. France had already moved first, banning E171 in food from January 1, 2020. Then Commission Regulation (EU) 2022/63 removed titanium dioxide from the EU's list of approved food additives entirely — it entered into force on February 7, 2022, with a six-month grace period, and the full ban took effect on August 7, 2022.

Now the other side of the file, quoted just as accurately.

Health Canada reviewed the same body of evidence and published its conclusion in 2022: there is "no conclusive scientific evidence that the food additive TiO₂ is a concern for human health." Its scientists argued that many of the studies driving concern used non-food-grade titanium dioxide, or artificially reduced particle sizes not representative of what people actually eat, and that food proteins can bind to the particles in ways that change their behavior in the gut. The UK Food Standards Agency and Food Standards Australia New Zealand reached broadly similar positions. Health Canada also said it "will continue to monitor the emerging science" and may revisit.

The FDA agrees with that camp — for now. It maintains that it "did not identify concerns related to potential genotoxicity based on the data available," and it pushed back specifically on EFSA's method, noting that some genotoxicity tests EFSA considered used "test materials not representative of the color additive" and "administration routes not relevant to human dietary exposure" — meaning particles injected rather than eaten.

But "for now" is doing real work in that sentence. In 2023, a coalition including the Environmental Defense Fund, the Center for Food Safety, and the Center for Science in the Public Interest filed a color additive petition (Docket FDA-2023-C-1487) asking the FDA to repeal 21 CFR 73.575 and pull titanium dioxide from food. The FDA opened it for public comment in the Federal Register on May 3, 2023. It remains under review. And the market has started moving without waiting for the ruling: Mars removed titanium dioxide from U.S. Skittles by the end of 2024 — a quiet fulfillment of a phase-out pledge it first made in 2016 — a change that only became widely known in 2025.

What is not known

This is the part where honesty matters most. The core scientific question — whether the titanium dioxide particles in the food you eat, at the doses you eat them, damage DNA in a way that matters for human health — is genuinely unresolved. Not "resolved in Europe's favor." Not "resolved in the FDA's favor." Open.

What is known: the compound is inert enough that most of what you ingest passes through and out. High-dose animal feeding studies did not show cancer. There is no strong human data showing that dietary titanium dioxide makes people sick.

What is not known: whether the nano-scale fraction accumulates in tissue over decades, whether it interacts with the gut lining or gut bacteria in ways that matter, and whether the standard toxicology tests are even measuring the right thing for a particle this small. EFSA said it couldn't rule the risk out. Health Canada said the evidence doesn't confirm it. Both statements can be true at the same time, and right now they are.

The pattern

Titanium dioxide is a case study in what happens when an approval outlives the science that justified it. A compound gets cleared in 1966 under the tools of 1966. The rule sets a limit and moves on. Fifty years later, a new field of measurement grows up around the exact thing that was waved through, and two competent agencies, looking at the identical evidence, split — one treating "we can't rule it out" as a reason to stop, the other treating "it isn't proven" as a reason to continue. The disagreement isn't a scandal. It's the honest shape of a question the original approval never had to answer.

We don't tell you it's poison. We don't tell you it's fine. Here is the file: a 1966 clearance, a 1% limit, an EFSA opinion that couldn't exclude genotoxicity, a Health Canada review that found no conclusive concern, an EU ban, an open FDA petition, and a candy company that already changed the recipe. You can read a label. Titanium dioxide is listed by name, or as E171, or as "color added." Now you know what it is when you see it, and you know that the world's food agencies don't agree on it either.

If you want the rest of the color-additive file — how these pigments get approved, which ones are already gone, and where else they turn up — that's what we build. Start with The Label Files, or browse the whole library.

Related reading: Red 40 — the dye also used to color fabric and five ingredients banned abroad but still in U.S. food.