The Colour on Your Mug

The mug you reached for this morning was almost certainly not chosen for being safe. It was chosen because it holds the right amount, because the handle sits well in the hand, or simply because the colour is cheerful at seven o'clock when very little else is. Most people own one like it: the favourite, the first one down off the shelf, with a band of bright colour running round the outside just below the lip.

That band is worth a second look, because it sits on the one part of a mug your lips really touch. The coffee inside meets your mouth for the half-second it takes to go past. The rim meets it far longer: drink after drink, morning after morning, your lower lip resting against the same narrow strip of the outer wall, the top two centimetres or so, from the lip down.¹ On a plain white mug that strip is nothing but glaze. On the favourite it is often where the decoration lives: a red, an orange, a gold line, a painted pattern that runs up to just short of the edge.

Colour like that has to be made from something, and the something is usually a metal. The bright reds and oranges are coloured with cadmium; lead turns up too, in older work. On its own that is not alarming: these are ordinary materials in a fired ceramic, and the inside of the cup, where the drink actually sits, is almost always clean.

The outside is where the colour goes, and it can get there two ways. On some mugs it is painted on and then sealed under a coat of clear glaze, shut behind a skin of glass. On others it is fired on afterwards, sitting on top of that skin, on the exact strip your lip drags across. And nothing about the mug in your hand tells you which method the cup you are holding used.

On-glaze colour sits on the outside of the glaze, not sealed under it

A glaze is glass. Before it goes into the kiln, a mug is porous clay, a cup you could not drink from, because the tea would soak straight into the wall. The glaze is a coat of powdered minerals brushed or dipped over that clay; in the heat of the kiln it melts, flows, and cools into a smooth skin of glass fused to the surface. That skin is what makes a mug a mug: waterproof, wipeable, hard enough to survive a thousand dishwasher cycles.

Colour can go on before that skin forms or after. Painted onto the raw clay and then covered with the clear glaze, a pattern ends up sealed beneath the glass, looking out through it the way a fish looks up through the surface of a pond. Fired onto the finished glaze afterwards, at a lower heat, colour becomes its own thin layer of coloured glass, fused to the outside.¹ The trade calls the first kind under-glaze and the second on-glaze, and the difference is not cosmetic: one colour is armoured behind glass, the other is exposed on the surface.

When the European Commission's own reference laboratory set out to study exactly this, its chemists had two coffee cups made specially: two little fifty-millilitre cups, one decorated over the glaze and one under it. Filled with a weak acid and left for a day, the cup with the colour on top gave up far more lead than the one with the colour sealed under the glass. The laboratory's own explanation was that the on-glaze decoration was "more sensible to the acidic attack": more easily got at, because it was sitting out where the acid could reach it.¹

So a single red band can be one of two quite different things: a colour behind glass, or a colour on the surface a mouth meets. Which of the two a given mug carries is not something you can tell by simply looking at the mug.

Ceramic poisonings are why the test exists, and the test only wets the inside of the cup

Why would anyone fill a mug with acid and measure what comes out of it? Because the metals that make a glaze do not always stay in it, and when they leave, they leave into whatever you are drinking.

The clearest cases are the frightening ones. In 2016 a woman in Toronto came to hospital twice with abdominal pain and a blood count that made no sense. Every test came back normal until a technician looked at a smear of her blood under the microscope and saw the red cells stippled with tiny dots: the classic fingerprint of lead. Her blood lead was thirty-six times the upper limit of normal.¹⁰ The source was a mug and a pot she had brought back from Mexico: the clay was clean, but the glaze coating them was seventeen per cent lead by weight, and it had been dissolving into her food.¹⁰ Her son, who used the mug less, was poisoned too, only less.¹⁰ Two years later, American doctors traced a pregnant woman's dangerously high blood lead to nothing in her home or her work, until they came to the ceramic mug she used for hot lemon water every day. The acid and the heat together had been breaking down its leaded glaze from the inside.¹¹

Poisonings like these are why the tests exist, and they are old news to the people who write food-contact rules. A 1970 paper in the New England Journal of Medicine was titled, plainly, "Earthenware containers as a source of fatal lead poisoning."¹² Every case that built the modern rules had the same shape: the danger sat on the inside, in the glaze that met the food and the drink.

So that is what the standard European test does. It fills the mug (to within a millimetre of the brim) with a weak acetic acid, roughly the chemistry of a sharp vinegary drink, and leaves it for a day. Then it measures the lead and cadmium that have migrated out of the glaze and into the liquid. Anything on the outside that a drink would never touch is masked off first, painted over so it cannot contribute.³ The test asks one careful question: if you keep an acidic food or drink in this vessel, how much metal ends up in it?

It is the right question for the inside of a mug. The trouble is that a mouth does not drink from the inside of the rim. It rests on the outside of it — the one strip the test paints over.

No EU law tests the 2 cm rim your mouth meets; only some countries measure it

The rule that governs mugs across the European Union is a short one, adopted in 1984 and largely unchanged since.³ It sets a limit for lead and cadmium released into food, and it checks compliance by filling the article. For any surface not meant to hold food, the method says to cover it with a protective layer before testing. So the outside of the mug, including the band a lip rests on, is deliberately kept out of the liquid.

The Commission's reference laboratory said as much, in a 2017 report on whether the rim ought to be tested too. "Currently," its chemists wrote, "there is no specific provision stipulated in the EU legislation for testing the migration from the rim."² That sentence is about European law only: no common European limit reaches the rim. It is not the same as saying nobody measures it.

Some countries measure it themselves. In Germany, the state food-safety laboratories check the two-centimetre drinking rim against a figure of two milligrams of lead that comes from a technical standard, DIN 51032 — a standard the same German authority is careful to note "is not a legal norm."⁵ The binding German law, like the European one, reaches only the fillable inside; the rim number is a benchmark the laboratories apply, anchored in the general duty that an article must not shed substances in harmful amounts. In Finland, a binding ministerial decree does name the lip rim directly.⁸ But it reaches only mugs made in Finland or imported straight from outside the EU, so a decorated cup that arrives lawfully from another European country slips past it.¹⁵ And from 29 May 2026, on a 2024 Benelux decision, the Netherlands began requiring makers to weigh the lead and cadmium in any part of a cup expected to meet food or the mouth, the outside included, when they judge a product safe. It is the first European rule to look past the fillable inside, though even it sets no fixed limit for the rim.^16,17

The curious thing is how the numbers agree. The German standard's two milligrams of lead per rim, and the identical two milligrams in an international standard used by testing laboratories, match down to the decimal, neither one a legal limit.^2,5 Everyone who bothers to measure the rim measures it against the same figure. Almost nobody is required to.^5,8

Rims are sometimes caught anyway. The Commission runs an alert system through which national inspectors flag dangerous goods to one another, and its records hold a long tail of decorated rims stopped at borders and pulled from shelves: a wine-glass rim that released 226 milligrams of lead, a glass cup giving up 163 milligrams from its rim.¹³ Almost all of those were drinking glasses, not ceramic mugs. Only a handful of the rim alerts concern ceramic drinkware at all: three consignments of Chinese mugs turned back at the Finnish border, one set of Chinese cups withdrawn from the German market, and, in 2010, a set of coffee cups detained in Finland that had been made in Germany.¹³ Where the rim shows up in the enforcement record, it is mostly a glass problem and mostly an import one — but not only.

All of which describes a surface that European law leaves to chance and individual countries watch unevenly. It does not describe what is actually on the rim of the ordinary mug in an ordinary kitchen: the kind nobody flags, nobody turns back, the kind that passes. For that, someone has to take a passing mug and measure the one band the rule skips.

The scientists who measured the rim found lead and cadmium on it, even though all below the limit

In 2025 a scientist at the University of Plymouth, Andrew Turner, did the plain thing nobody was obliged to do. He took ordinary ceramic mugs (new ones and second-hand ones) and, instead of only filling them, measured what came off the lip area, the outer band a mouth meets. From the interiors, very little: well within the European limits, as the interiors of mugs almost always are. From the lip areas, more: as much as 934 micrograms of lead and 115 of cadmium per litre of acid over a day.⁴ The largest amounts came, in his words, from mugs "where the rim area was overglazed with decorative Cd sulphoselenide pigments", where a red or an orange had been fired onto the outside of the rim.⁴ Lip-area migration, he wrote, is "not currently regulated in the EU but appear[s] to result in greater exposures that warrant further consideration and investigation."⁴

A decade earlier, a chemist at the Bavarian state laboratory, Martin Heimrich, had already turned to the same band, and to the most seasonal of objects. In November and December of 2011 he took twenty-six decorated Christmas mugs (the bright painted kind sold for mulled wine), turned them upside down and dipped their rims in acid. The result was, in his own word, welcome: every one of the twenty-six sat many times below the limit, and the liquid the mugs had been filled with held no detectable lead or cadmium at all. And yet the rims were not blank. Lead came off nine of the twenty-six rims and cadmium off eight: nowhere near enough to fail, but there, detectable, on the strip a drinker's lip would cross.⁵ His laboratory's wider survey work turned up something stranger still on some blue-decorated rims: cobalt coming off at well over 100 micrograms per rim, a metal for which no limit exists anywhere to measure it against, and the authority said, in its own report, that one ought to.⁶

The pattern held when others looked. In a survey run in 2022 and written up the following year, two chemists at the state laboratory in Baden-Württemberg, Magdalena Köhler and Natalie Stark, tested twenty ceramic items (dinner sets, children's beakers, teapots) and found that every single one met the law. They passed. But the two analysts added a sentence that sat oddly beside their own clean results: the limits every sample had comfortably met were set in 1984, and no longer match what toxicology now understands about these metals.⁷ Heimrich's laboratory had found the same doubleness across its routine work (most ceramic ware complies, twenty-one items over the limit in one year and only a single one the next), and yet a measurable trace of lead or cadmium still shows up in as many as forty per cent of samples, all of them within the law. It shows up, the laboratory noted, above all at the drinking rim.⁶

So the surface the European rule skips turns out not to be a place where anything alarming is happening. The people who looked found metal on the rim, and found almost all of it comfortably legal. Which raises the question a passing grade is supposed to settle: if the mug meets the limit, why should any of this matter?

The limit is a loose 1984 number, and no one has measured what a mouth takes off a rim

Two things sit inside a passing grade, and they pull in opposite directions.

The first is that the limit is old and generous. The number a mug's interior is measured against (four milligrams of lead per litre) was fixed in 1984, and the scientists who have looked at it since think it far too loose.³ The Commission's own laboratory has drawn up a revised figure that would cut the allowable lead by a factor of four hundred, down to the level tolerated in drinking water. That work was finished in 2017.² The 1984 number is still the one in force. So "below the limit" means below a line drawn more than forty years ago, by people who knew less than is known now, and which its own custodians have proposed to move by a factor of hundreds.

The second thing cuts the other way, and it belongs in the same breath. When the Commission's chemists tested their two purpose-made cups, the worst of them (the one with the colour fired on top of the glaze) gave up about 1.24 milligrams of lead from its rim area over a day in acid.¹ That sits below the two-milligram figure the German and international standards use for the rim, a benchmark rather than a law.^2,5 The most exposed cup in the whole experiment, soaked in acid for twenty-four hours, still came in under the rim number. Almost everything does.

And here is where the passing grades stop reassuring and start being beside the point. Every one of these numbers (Turner's, Heimrich's, the Commission's) comes from a mug filled or dipped and left in acid for a day. None of them is what a mouth does. A day's immersion is not a sip, and no study we could find has measured a sip. No published measurement was located for what saliva lifts off a decorated rim, the wetted area of a lip, the seconds of contact, or the number of times a day a mouth returns to the same warm painted band, over years. The chain that runs from "there is cadmium on this rim" to "this much cadmium enters this person" has a link missing in the middle, and no one has yet supplied it.

So neither comfortable answer survives. The metal is really there, on the surface a mouth meets. Almost every mug passes the test. And the test was never designed to answer the question a drinker would actually ask. Which leaves one thing a person can act on at a shelf: whether there is a way to choose the mug with less on its rim to give up — and to see it before buying.

The label never says which side of the glaze the colour is on; the rim's surface often shows it

The mug itself gives very little away. It may carry a word or two about safety ("lead free," most often), but that phrase is about content, not position. In the trade it means the colour holds no more than a trace of lead by weight, and it can sit on a range that still includes cadmium reds and oranges.⁹ It says nothing about whether the colour is under the glaze or on top of it, which is the thing that governs how much can come off.

There is a document that records how a mug was made and tested: a declaration of compliance, which in England every ceramic maker or seller must send along with the article all the way to the shop counter. But it is not written for the shopper. The law gives the right to demand the test results behind it to exactly two kinds of body (the local food authority and the port health authority), and to no member of the public at all.¹⁴ And for a second-hand mug (the inherited one, the charity-shop find), the duty does not apply: there is no declaration to ask for, and no one who made it is still in the chain.¹⁴

The colour itself is where the material truth sits, and it is oddly specific. Run down a modern decorating supplier's colour chart and most of the palette (the greens, the blues, the paler yellows) is free of both lead and cadmium. The industry has spent decades moving that way, and a 2023 colour chart carries twenty-three such colours.⁹ But at the hot end of the spectrum (the bright reds and oranges, the deepest yellow), cadmium is still there, in four colours the chart footnotes and warns will mix with nothing else.⁹ That is why the band that catches the eye at the shelf, the cheerful red ring, the orange flowers, is the one most likely to carry the metal. The colour maker, for its part, prints a caution in its own technical sheet: the release must be tested by whoever fires the mug, "for all products manufactured under his technical production conditions."⁹ The powder is controlled; the finished object is somebody else's responsibility.

What is left is what a person can actually see. A rim with no decoration on it has removed the one thing most likely to migrate: the coloured layer fired onto the strip a lip meets. A plain glaze is not nothing, but it is the smaller worry. A colour that runs under a glassy sheen, looking out through it, is sealed behind the same barrier that keeps the coffee off the clay. A red band that sits proud of the surface, matte where the rest of the mug shines, is likely the on-glaze kind, the exposed kind, on exactly the two centimetres that matter. It is a strong hint, not a certain test. None of this needs a laboratory; it needs a look at the rim in the photograph before buying, and a preference for the mug whose colour is under the glass, or whose rim simply carries no colour at all.

It is a modest kind of control, and it stops where the knowledge stops. You can see which side of the glaze the colour is on, and choose the sealed rim or the plain one over the painted band. What no one can yet tell you (what no study has measured) is how much of that band a real mouth takes, sip after sip, morning after morning, from the one part of a mug the rules were never written to watch.