REACH vs TSCA: Why EU and US Chemical Rules Differ

If you read enough of this site, you start to notice that almost every individual story (the dye, the flour additive, the shampoo preservative, the soda emulsifier) is really the same story wearing different clothes. Europe restricts something the US permits, and the explanation always bottoms out in the same place: the two systems made opposite choices about a single question.

The question is: before a chemical can be sold, who has to prove what?

In Europe, the answer is “the company, and it has to prove safety first.” In the United States, the answer is, roughly, “the government, and only after there’s a problem.” Those two answers are written into two laws, REACH in the EU and TSCA in the US, and once you understand how they differ, you can predict the outcome of nearly every “banned in Europe” debate before reading the details.

This is the explainer I wish existed when I started. Nobody writes it for normal people; the only thorough sources are compliance-firm white papers aimed at corporate clients. So here it is, in plain English.

The two laws, in one sentence each

REACH (Registration, Evaluation, Authorisation and Restriction of Chemicals, formally Regulation (EC) No 1907/2006) is the EU’s chemicals law, adopted in 2007 and run by the European Chemicals Agency (ECHA) in Helsinki. Its governing idea is “no data, no market”: if you want to manufacture or import a chemical into the EU above one tonne per year, you must register it first, filing a full safety dossier. No dossier, no sale. (EU-OSHA; Ecomundo.)

TSCA (the Toxic Substances Control Act of 1976, run by the EPA) is the US chemicals law. Its original governing idea was essentially the reverse: existing chemicals are presumed safe, and the EPA can restrict one only after assembling evidence of “unreasonable risk.” The burden sat with the regulator, not the manufacturer. (comparative analysis, Taylor & Francis.)

That’s the whole divergence in miniature. REACH puts the burden of proof on industry, before market entry. TSCA put it on the government, after market entry. Everything else follows.

How REACH actually works

The mechanism is worth understanding because the “burden on industry” framing is abstract until you see the machinery.

Registration. A company wanting to sell a chemical in the EU compiles a registration dossier: identity of the substance, its hazard properties, how it’ll be used, exposure scenarios, and a risk-management plan. The more of the substance you sell, the more data you must supply. This dossier goes to ECHA before the product can legally be placed on the market. If you don’t register, you can’t sell. That’s the literal meaning of “no data, no market.” (Ecomundo.)

Evaluation. ECHA and member states review dossiers, can demand more testing, and scrutinize substances of concern.

Authorisation and Restriction. The most hazardous substances (carcinogens, mutagens, reproductive toxicants, persistent bioaccumulative chemicals, endocrine disruptors) get flagged as “Substances of Very High Concern” (SVHCs). These can be placed on a list that eventually requires specific authorisation to use at all, pushing the market toward safer substitutes. Restriction can limit or ban specific uses outright.

The key cultural feature: REACH shifted the burden of proof from authorities to companies, which now have to demonstrate the safety of manufacturing and using their substances. (ResearchGate, REACH and precaution.) A chemical with an incomplete safety story doesn’t get the benefit of the doubt. It gets kept off the shelf.

How TSCA worked, and the grandfather clause that broke it

TSCA’s original design had a flaw so consequential it shaped the entire US chemical landscape for forty years.

When TSCA passed in 1976, roughly 62,000 chemicals were already in commerce. The law grandfathered all of them in: they were treated as acceptable without any safety review, simply because they already existed. New chemicals got a light-touch pre-market notice, but the EPA’s power to actually restrict any chemical, old or new, required clearing a famously high legal bar: proving “unreasonable risk” while choosing the “least burdensome” possible regulation.

That bar turned out to be almost insurmountable. The defining moment came in 1991, when a federal court overturned the EPA’s attempt to ban asbestos (asbestos, a substance whose carcinogenicity was not in scientific doubt), ruling the agency hadn’t adequately met TSCA’s least-burdensome standard. If the EPA couldn’t ban asbestos under TSCA, it effectively couldn’t ban much of anything. In its first four decades, the agency restricted only a handful of chemicals.

So while REACH’s default was “prove it’s safe or it’s off the market,” TSCA’s default was “it’s on the market, and good luck proving it shouldn’t be.” Under TSCA, existing toxic substances were presumed safe until the EPA had conclusive evidence otherwise. (Taylor & Francis comparison.)

The 2016 reform that helped but didn’t converge

In 2016, Congress passed the Frank R. Lautenberg Chemical Safety for the 21st Century Act, the first serious update to TSCA in forty years. It was a real improvement: it required the EPA to systematically evaluate existing chemicals, removed the crippling “least burdensome” language that sank the asbestos ban, and mandated protection of vulnerable and highly-exposed populations. (NIH PMC, Lautenberg analysis.)

But, and this is the part that matters for everything on this site, the Lautenberg amendments did not adopt the EU’s core move. The US did not put the burden of generating safety data, doing the risk assessment, and managing the risk onto industry the way REACH does. (NIH PMC.) The EPA still drives the process. It still works through the existing inventory chemical by chemical, slowly. The default never flipped from “permitted” to “prove it.” TSCA got better at its job; it didn’t become REACH.

That’s why “the US updated its chemical law in 2016” and “the US still permits things Europe bans” are both true at once. The reform fixed the machinery without changing the default setting.

The precautionary principle, fairly described

You’ll often hear the EU approach summarized as “the precautionary principle” and the US approach as “risk-based” or “sound science.” Both labels are loaded, so let me describe the actual difference without the cheerleading.

The precautionary posture says: when there’s credible evidence of potential harm but the science isn’t settled, lean toward restriction. Uncertainty counts as a reason for caution. The cost is that you’ll sometimes restrict things that later turn out to be fine (false positives), and you may keep useful chemicals off the market.

The risk-based posture says: restrict only when there’s a demonstrated, quantified risk. Uncertainty alone isn’t enough to act. The cost is that you’ll sometimes permit things that later turn out to be harmful (false negatives), and harmful substances can stay in use for a long time while evidence accumulates (see brominated vegetable oil, where that “long time” was 54 years).

Neither is objectively correct. They’re different ways of distributing the same unavoidable error. A precautionary system over-restricts; a risk-based system under-restricts. Which mistake you’d rather make is a values question, not a scientific one, and pretending otherwise is how both the “Europe is enlightened” crowd and the “Europe is anti-innovation” crowd get the story wrong.

There’s even a case where the EU’s faster machinery helps American consumers lose out: European sunscreens contain modern UV filters approved years ago in the EU but still not cleared in the US, because the FDA regulates them as drugs requiring a high evidentiary bar. Same structural pattern, opposite direction: the system that’s better at restricting questionable chemicals is also better at approving good ones. We get into that in Why do I feel better in Europe?.

A note on food, cosmetics, and where these laws actually bite

One clarification, because it confuses people: REACH and TSCA are industrial chemicals laws. Food additives and cosmetics each have their own separate regimes: the EU’s food additive regulation (1333/2008), its Cosmetics Regulation (1223/2009), the US food-additive and GRAS system under the FDA. So strictly speaking, your titanium dioxide or parabens question is decided under those laws, not REACH or TSCA directly.

But the philosophy is shared across all of them. The EU’s food and cosmetics rules carry the same “prove it’s safe, and uncertainty defaults to restriction” logic that REACH encodes for industrial chemicals, which is why the EU cosmetics regulation has banned or restricted well over a thousand substances while the US has restricted a far smaller number. (formaldehyde is a good worked example.) REACH and TSCA are the clearest, most explicit statements of the two philosophies, which is why they’re the right lens. The same fork in the road runs through every regulatory domain.

So which system is “better”?

The honest answer is that it depends on which error you’re more afraid of, and that reasonable people weigh that differently.

If you think the bigger danger is harmful chemicals sitting in products for decades while regulators slowly build a case, the EU model is plainly better, and the US track record (asbestos, BVO, the GRAS loophole) gives that view a lot of ammunition. If you think the bigger danger is a precautionary system restricting useful, ultimately-safe substances on incomplete evidence and raising costs, the US model has a defense, though in practice the US’s problem has been under-action, not over-caution.

What I’d push back on, hard, is the lazy version in either direction: “Europe bans everything because it’s scared of science” and “America poisons you because it doesn’t care” are both cartoons. The truth is duller and more useful: two systems made a defensible, opposite choice about how to handle uncertainty before a product reaches you, and that single choice, written into REACH and TSCA in the 1970s and 2000s, is the engine generating nearly every story on this site.

The takeaway

REACH and TSCA answer the same question (may this chemical be sold?) with opposite defaults. REACH says “no data, no market”: industry must prove safety before market entry, and uncertainty defaults to restriction. TSCA presumed existing chemicals safe and made the government prove harm afterward, a bar so high the EPA couldn’t even ban asbestos until the 2016 Lautenberg reform partly fixed the machinery, without ever flipping the default. The result is a precautionary EU and a risk-based US, each making the opposite mistake: Europe sometimes over-restricts, America routinely under-restricts. Which is “better” is a values question about which error you’d rather live with, and almost every “banned in Europe” headline is just this one structural difference, applied to a specific molecule.