26 Food Additives Banned in Europe but Legal in the US (2026)

Walk down a supermarket aisle in Brussels and one in Boston and you are looking at two different chemical realities. The packets may carry the same logo, the same colours, the same cheerful promises on the front. Turn them over and the ingredient lists diverge — sometimes by a single line, sometimes by five or six.

This is not an accident of taste or supply chains. It is the direct result of two regulatory systems that start from opposite assumptions about who has to prove what. As of 2026, at least 26 substances that the European Union has banned, restricted, or simply refused to approve remain legal and in use across the American food supply.

This guide lists them, explains the specific scientific objection behind each one, and — more usefully — explains the structural reason the two lists will probably never converge. It also covers what changed in 2025 and 2026, because this is a moving target: the United States banned Red 3 from food in January 2025 with a compliance deadline of January 2027, and pulled brominated vegetable oil in August 2024. The gap is narrowing in places. In others it is not moving at all.

Before the list, the mechanism. Almost every specific disagreement below traces back to a single structural difference in how the two systems handle uncertainty.

The European Union operates on what is formally called the precautionary principle. A substance must be demonstrated safe before it enters the food supply. When new evidence raises doubt about something already approved, the European Food Safety Authority re-opens the file, and the burden stays on the substance to justify its continued presence. If the evidence is ambiguous, ambiguity counts against approval. Authorisations are reviewed on a schedule rather than left permanently open.

The United States runs a different system, and the pivotal piece of it is a category called GRAS — Generally Recognized As Safe. A substance in this category does not require pre-market approval from the Food and Drug Administration at all. A manufacturer may conclude on its own that an ingredient is safe, based on its own assembled evidence, and begin using it. Notifying the FDA is voluntary. Once something is in use, it generally stays in use unless the agency takes affirmative action to remove it — and that action requires building an evidentiary case, surviving industry comment periods, and often litigation.

The practical consequence is asymmetric inertia. In Europe, doubt tends to remove a substance. In America, doubt tends to leave it in place. Two regulators can read the same study and reach opposite operational conclusions without either behaving irrationally, because they are answering different questions. The European question is "has this been shown safe enough to keep?" The American question is "has this been shown harmful enough to remove?"

This matters for how you read the list below. A substance appearing on it does not automatically mean American regulators examined the evidence and declared it harmless. In many cases it means no one has yet assembled the case required to force it out.

There is movement on this. The FDA's 2026 priority agenda includes a proposed regulation that would require companies to submit GRAS notices to the agency for all new substances claimed as GRAS, rather than self-affirming in private. It also includes post-market safety reviews of specific substances — phthalates, propylparaben, BHA and BHT among them. If the GRAS proposal becomes a final rule, it closes the single largest structural gap between the two systems. As of this writing it is a proposal, not law.

Each row below reflects the regulatory status as of 2026. "Restricted" means permitted only in specific categories or below specific limits rather than banned outright.

Notice how many entries in the US column now read "under FDA post-market review in 2026." That is the single most important pattern in this table. Four of these substances are actively being re-examined under the FDA's 2026 Human Foods Program agenda. This list is likely to be shorter in two years than it is today.

If you want one substance that illustrates the whole disagreement, this is it.

Titanium dioxide is a whitening and opacifying agent. It makes icing brilliantly white, gives chewing gum its opaque shell, brightens coffee creamers and sauces, and appears in the coatings of sweets and supplements. It has no flavour and no nutritional function whatsoever. Its entire purpose is visual.

In May 2021, EFSA published an opinion concluding that titanium dioxide could no longer be considered safe as a food additive. The specific reasoning matters: the panel could not rule out genotoxicity — the potential to damage genetic material — and, critically, could not establish a safe daily intake level. When a regulator cannot identify any dose it is confident about, the precautionary framework has only one available answer. The EU ban took effect in 2022.

In the United States, titanium dioxide remains permitted in food at levels up to one percent by weight. The FDA's position has been that the available evidence does not establish harm at the levels consumed. Note again the shape of that sentence. It is not a finding of safety; it is a finding that harm has not been demonstrated to the standard required for removal.

For a shopper this creates a genuinely strange situation. A confectionery brand selling in both markets will often reformulate for Europe and keep the original formula for America. Same brand, same packet design, different ingredient list — and the European version exists purely because a regulator demanded it.

What to look for on a label: "titanium dioxide", "E171", or "colour added". In the US it may simply appear as "artificial colour", which is why the ingredient list alone does not always tell you.

The Red 3 story is the most quotable example of regulatory inertia in the entire American system, and it is worth telling precisely.

Erythrosine, known as FD&C Red No. 3, is a synthetic cherry-red dye. In 1990, the FDA banned it from cosmetics and topical drugs. The reason was studies showing thyroid tumours in male rats at high doses, which triggered the Delaney Clause — a provision in US law stating that no additive shown to induce cancer in humans or animals may be deemed safe.

Here is the part that does not make sense on first reading. The same agency, applying the same clause, left Red 3 legal in food. For thirty-five years you could not legally put this dye on your lips, but you could legally eat it in sweets, baked goods, and maraschino cherries.

That ended in January 2025, when the FDA finally revoked authorisation for Red 3 in food. Manufacturers have until January 2027 to reformulate food products, and January 2028 for ingested drugs. Meanwhile the EU had restricted erythrosine for decades, permitting it only in narrow applications such as certain cocktail cherries.

The lesson is not that Red 3 was uniquely dangerous. It is that the American system can hold a known, formally recognised contradiction in place for three and a half decades, because removing something requires affirmative effort and leaving it requires none.

That deadline is live right now. Between now and January 2027, thousands of American products are being reformulated. If you are reading an ingredient list in the US today and see Red 3, you are looking at a product on borrowed time.

Potassium bromate is a flour treatment agent. It strengthens dough, produces higher-rising loaves and a finer crumb, and speeds up industrial baking. In theory it converts entirely to harmless potassium bromide during baking.

In practice, that conversion is incomplete when baking time, temperature, or the quantity used varies — and residual bromate remains in the finished bread. The International Agency for Research on Cancer classifies potassium bromate as possibly carcinogenic to humans. The EU does not permit it. Neither do the United Kingdom, Canada, Brazil, China, or Argentina.

It remains legal in the United States federally. California passed state legislation to ban it, which is itself a telling development: when the federal system does not move, American food regulation increasingly happens at state level, producing a patchwork where the same product may be legal in one state and not another.

This is one of the harder substances to detect as a consumer, because bromate itself is not what appears on the label. Look for "potassium bromate" or "bromated flour" in the ingredient list. Many American brands have voluntarily removed it and say so on the packet.

Balance matters here, and most articles on this topic omit it entirely.

The EU does not win every comparison. There are areas where American rules are equal or tighter, and pretending otherwise damages the credibility of the whole exercise.

The United States has mandated added-sugar declaration on the Nutrition Facts panel since 2016 — a line item showing how much sugar was added during processing rather than occurring naturally. The EU does not require this; its nutrition declaration lists total sugars only. Australia formally abandoned a proposal to require it in March 2026. On this specific point, the American label tells you more.

American allergen labelling, under FALCPA, requires plain-language declaration of major allergens, and sesame was added as the ninth major allergen in 2023. The EU requires allergen emphasis within the ingredient list, which some researchers consider harder to scan quickly.

The United States also sets a zero-tolerance standard for Listeria monocytogenes in ready-to-eat foods, which is stricter than the EU's tolerance threshold in certain product categories.

The honest summary is that Europe is more restrictive on chemical additives, particularly colours and preservatives, while the United States is stronger on certain labelling disclosures and on some microbiological limits. Neither system is comprehensively safer.

One of the most consequential developments in this area is not happening in Washington at all.

When a federal system moves slowly, the pressure does not disappear — it relocates. Individual American states have begun passing their own additive bans, producing something the American food industry has spent decades trying to avoid: a patchwork where the legal formulation of a product depends on which state it is sold in.

California moved first and most visibly, passing legislation banning several additives including potassium bromate, brominated vegetable oil, propylparaben, and Red 3 — the last of these ahead of the FDA's own action. Other states have introduced similar bills. The pattern is that state legislatures, facing fewer procedural hurdles than a federal agency and often more direct constituent pressure, can act on a timescale of months rather than decades.

For manufacturers this creates a straightforward commercial calculation. Producing two versions of a product for different states is expensive and logistically awkward. Producing one compliant version for the whole country is simpler. In practice, a large state acting alone tends to become a de facto national standard — which is precisely why industry groups have historically lobbied for federal pre-emption of state food laws.

For consumers the implication is subtler but useful: American product formulations may change without any federal announcement, driven by state law you never read about. If a brand quietly reformulates, that is often why.

It also means the list in this article is genuinely a moving target within the United States itself, not only between countries. A substance can be federally legal and simultaneously unlawful in the state where you are standing.

Abstract lists are less useful than knowing which shelf to be careful on. Here is where the divergent additives concentrate, by category.

The last two rows deserve a specific warning. Ractopamine and rBGH are production inputs, not additives mixed into the final product, which means they generally do not appear on an ingredient list at all. No amount of careful label reading will reveal them. The only signals available to a consumer are voluntary claims such as "no added hormones" or certified organic status, where the certification standard itself prohibits them.

This is an important limit on the label-reading method. Ingredient lists tell you what was added to the food. They do not tell you what was done to the animal or the crop before it became food.

Three specific things are in motion, with dates attached.

The GRAS reform proposal is the largest. The FDA's 2026 agenda commits to publishing a proposed regulation requiring submission of GRAS notices to the agency for all new substances claimed as GRAS. This would end self-affirmation in private — the single mechanism that allows substances to enter the American food supply without regulatory review. Watch whether it survives the comment period intact.

Post-market reviews of BHA, BHT, propylparaben and phthalates are scheduled for 2026 under the FDA's newly built Office of Post-Market Assessment. All four appear on the divergence list above. If those reviews conclude against the substances, four rows disappear.

The Red 3 compliance deadline of January 2027 is the nearest and most certain. Reformulation is happening now across thousands of products.

Moving in the other direction, the EU published revised guidance on 20 January 2026 setting the scientific data requirements for food additive applications, applying to all applications submitted from 20 July 2026. And in January 2026 the EU introduced new restrictions on several thickeners and stabilisers specifically in foods for infants, young children, and patients requiring foods for special medical purposes — tightening rules for the most vulnerable consumers rather than the general population.

The direction of travel is convergence, slowly, with America moving toward the European position rather than the reverse. But "slowly" is doing real work in that sentence. Titanium dioxide has been banned in the EU since 2022 and there is no active US process to remove it.

The most useful thing to take from this is not the list. Lists go stale, and this one will look different in two years.

The useful thing is the mechanism. When you see that a substance is permitted in one country and not another, the difference usually is not that one regulator found it safe and the other found it dangerous. The difference is usually that one system requires proof of safety to stay in, and the other requires proof of harm to be removed. Those two rules produce different outcomes from identical evidence.

That reframes the practical question. Instead of asking "is this banned somewhere?", ask "is this here because it was tested and cleared, or because no one has yet made the case to remove it?" For a substantial portion of the American additive supply, the honest answer is the second.

Reading the back of the packet is the only reliable way to close that gap yourself, and it takes about forty seconds once you know which part of the list matters.