Is your teaware safe? Lead, adulterated clay and standards
The risk lies in the layer that touches water and lips: lead and cadmium in glaze and overglaze, worst with hot, acidic liquids. Unglazed zisha from pure ore released no lead in a sample of 32 pieces; dyed clay is the worry. Limits differ several-fold between countries.
- Main risk:
- lead and cadmium from glaze and overglaze
- Cup, China and ISO:
- lead up to 0.5 mg/L in 24 h
- Cup, EU directive:
- lead up to 4.0 mg/L
- Pure-ore zisha:
- Pb and Cd below the quantification limit
- Dyed zisha:
- barium, manganese, cobalt, chromium
- Glass thermal shock:
- ~55 °C household, ~183 °C borosilicate
Safety standards don't test the object as a whole, only what migrates from it into the liquid, so the layer that touches water and lips is what counts. A porcelain cup with underglaze decoration is safer than one with the same picture painted over the glaze, and a Yixing pot made from pure ore is a different thing from one made of dyed clay.
Risks from the tea itself are covered in safety in the cup, and pores and heat in teaware materials. The question here is narrower: what can leach out of the ware.
Glaze and decoration: lead and cadmium
Paint under the glaze is sealed in. Underglaze and in-glaze decoration lies beneath a layer fired at 1200–1300 °C, and the wall of such a cup feels smooth. Painting on top of the glaze is another matter. The overglaze enameloverglaze enamelA pattern painted over finished glaze and fixed in a second, cooler firing. It sits on the surface, so it wears off and can release metal. (釉上彩, Japanese 上絵) is fired below 900 °C, and a lead flux lets the paint melt at about 800 °C and come out clear. That is where the lead and cadmium come from.
You can tell the techniques apart by touch: raised relief and a different sheen give away overglaze. Japanese Kutani, with its five colors, historically relied on lead paints, and the Ishikawa prefectural institute later developed a lead-free palette for local workshops.
The rim is a special case. In a sample of 14 tumblers decorated near the edge, 11 released more than 3 mg of lead, though the figure survives only in a secondary summary of a Danish memorandum. The cautious conclusion is to keep decoration on the outside, away from the lips and off the inner wall.
The worst case is craft pottery with lead glaze. In a 2016 Canadian clinical case, the glaze on a cup and a pot from Mexico contained up to 17% lead by weight, and the patient's blood lead reached 3.6 µmol/L. In Hungary in 1998, an 18-year-old ceramics apprentice got lead poisoning from tea with lemon she kept in her own lead-glazed teapot. These are isolated reported cases, not statistics, but they share a mechanism: acidic liquid and long contact.
"Chemical" zisha
Unglazed Yixing clay on its own passes the tests. Chinese researchers measured the release of 24 elements from 32 Yixing pieces: teapots, boiling pots, pots and electric stewers. Lead, cadmium, chromium, nickel and cobalt all fell below the limit of quantification. Teapots released the most aluminum, magnesium and barium; boiling pots released mostly aluminum and magnesium.1 For the clay itself, see Yixing clays.
The trouble is adulterated clay. Mining at Huanglongshan (黄龙山) was restricted, and according to a 2010 CCTV program, workshops brought in clay from Zhejiang and Anhui and dyed it. Of 15 pieces tested, 13 showed abnormal release of barium, manganese, cobalt and chromium; only two of the 15 were natural ore. This comes from press accounts of the broadcast, not a scientific paper.
The scientific record is more cautious. Spectral analysis of 1,085 pieces found an average of 0.0295% barium in ore (109 samples), against about 0.380–0.398% in clay with added barium carbonate. Dated pieces from 1979 onward held 0.2388–0.3877% barium; the authors link that year to the start of mass addition, but there is no independent dating.2 The method, X-ray fluorescence, leaves the object intact. You can't spot a chemical pot by eye.
How dangerous this is remains unknown. The Chinese ceramics standard GB 4806.4 sets no barium limit, and the samples differ: pieces bought at random for a TV program versus a market sample in a lab study. Dyed clay is a reason for caution, but none of the sources found prove that tea brewed in such pots does harm.
Wax instead of skin
New zisha is sometimes waxed to hide poor clay and add shine. The test is simple: pour boiling water over the outside. If the water beads and rolls off, and afterward a white film shows up that won't wash away, the surface is waxed. To remove the wax, boil it off: fill with cold water, bring to a boil, boil for 10 minutes, let it cool naturally, wipe, and repeat three times. The collector who gives this advice advises against doing it to zhuni (朱泥).
Metal, glass and the electric kettle
"Pewter." Many "pewter" (锡) vessels turn out to be tin-lead alloys. Zhejiang's provincial disease-control center recorded 12 lead poisoning outbreaks in 2011–2022, with 40 people ill and 20 hospitalized. In 11 of them the cause was yellow wine warmed in "pewter" jugs; in one, water from a zisha teapot and a kettle was suspected. The US Food Code allows no more than 0.05% lead in a food-contact alloy and 0.2% in solder. Pewter caddies are covered under tea caddies.
Brass, solder, silver plating. In Brussels in 2000, a family was poisoned by lead from a traditional metal teapot for mint tea, and those teapots were pulled from sale. Belgian researchers later sampled such teapots: most released a lot of lead, and it came from the solder at the spout, not from the tea. Silver plating has its own problem. The US FDA detains silver-plated ware when extractable lead averages 7 µg/mL or more for adult items and 0.5 µg/mL or more for children's. Other metals are covered in metal teaware.
The electric kettle. The hazard is an exposed nickel-plated element. In a 1998 German study of 8 used kettles, one exceeded 50 µg/L of nickel and two exceeded 20 µg/L, the EU limit for drinking water. After descaling with citric acid, release rose as much as 50-fold (peaking at 640 µg/L), and it took five boils to get back below 20 µg/L. After cleaning, pour out the water from the first five boils. A closed heating plate avoids the problem; see kettles.
Glass. Here the risk is mechanical. Ordinary soda-lime glass survives a sudden temperature jump of about 55 °C, and borosilicate glassborosilicate glassGlass with added boron oxide. It expands less on heating than ordinary soda-lime glass and survives a much bigger temperature jump. about 183 °C. Those figures come from a simple model by American ceramists; other authors, allowing for heat transfer, got 80 and 270 °C, but the order of magnitude holds. Partial tempering gives household glass little extra margin. Pour boiling water into a cold vessel of ordinary glass and it will almost certainly crack; see glass teaware.
Standards: every country draws its own line
The method is the same everywhere: fill the ware with 4% acetic acid for 24 hours at 22 °C and measure the metal releasemetal releaseHow much metal migrates from the ware into the liquid. Standards measure it in milligrams per liter after 24 hours in 4% acetic acid.. Only the limits differ.
| What is tested | China, GB 4806.4-2016 | ISO 6486-2:1999 | US, FDA | EU, 84/500/EEC |
|---|---|---|---|---|
| Cups, lead / cadmium | 0.5 / 0.25 mg/L | 0.5 / 0.25 mg/L | 0.5 µg/mL (lead) | 4.0 / 0.3 mg/L |
| Small hollow ware, under 1.1 L (teapot) | 2.0 / 0.30 mg/L | 2.0 / 0.5 mg/L | 2.0 µg/mL (lead) | 4.0 / 0.3 mg/L |
| Cooking ware | 3.0 / 0.30 mg/L | 0.5 / 0.05 mg/L | – | 1.5 / 0.1 mg/L |
1 µg/mL equals 1 mg/L. For a cup, the Chinese and ISO limits are eight times stricter than the European one: 0.5 against 4.0 mg/L of lead. The same object can be legal on the EU market and banned in the US. The FDA has no separate teapot category and classes teapots as small or large hollow ware by volume.
Release also varies within a batch, by up to 60%, so the true average should sit well below the limit. ISO tests four samples, the FDA six.
The EU rules are changing. The European Commission is preparing lower limits for lead and cadmium and plans to extend the rules to glass and enamel; its working papers call cuts of about 400-fold for lead and 60-fold for cadmium justified. The sources found give no final limits. The Danish institute DTU proposed 1.5 µg/L for lead, but that is a proposal, not a binding limit.
China also has a separate standard for zisha ware, GB/T 10816-2024, in force since July 1, 2025. It caps water absorption at 6.0% and sets thermal resistance for a 180 → 20 °C change. Compared with the 2008 edition, it drops the lead and cadmium release requirements, and it is a voluntary quality standard. Safety limits for ceramics come from the mandatory GB 4806.4.
How to check at home, and what to buy
There is no home test for lead in glaze. The EPA recognizes three kits (LeadCheck, D-Lead, the Massachusetts kit), but only for paint on wood, ferrous metal, plaster and drywall, and none passed both accuracy criteria.3 Nobody has validated the strips for glaze or zisha. Only a lab test by the standard's method is reliable: 4% acid, 22 °C, 24 hours, atomic absorption or ICP analysis.
That leaves careful looking and careful buying.
- Run a finger over the pattern on a new decorated cup. Relief and a different sheen point to overglaze. Replace cups with worn decoration.
- Chinese food-contact ware should cite GB 4806.4 (ceramic) or GB 4806.9 (metal) along with the factory name. The Suzhou market authority advises against buying ceramics without a factory name and standard number, and suggests boiling new pieces in water, with a drop of vinegar if you like, before first use.
- Keep acidic drinks out of doubtful ware. Tea with lemon, hibiscus blends and cider should not sit for hours in glazed ware of unknown origin. The type of tea barely affects metal release; what matters is acidity, temperature and time.
- Don't leave brewed tea standing in "pewter", brass or silver-plated vessels. A lead alloy leaches even without heat.
- After descaling an electric kettle, boil it five times and pour the water out.
- Test new zisha with boiling water on the outside, as described above.
If you adopt one simple habit, make it drinking from white porcelain with no decoration at the rim, so the paint never touches your lips.
Common questions
Does a Yixing zisha teapot leach lead?
Pure ore did not, in the one study found. In 32 Yixing pieces, lead, cadmium, chromium, nickel and cobalt in 4% acetic acid all fell below the limit of quantification. The worry is added dyes and fillers: a 2010 TV investigation found abnormal release of barium, manganese, cobalt and chromium in 13 of 15 pieces tested.
Can I test teaware for lead at home?
No. The US EPA recognizes three kits, but only for paint on wood, metal, plaster and drywall, and none met both accuracy criteria. Nobody has validated the strips for glaze or zisha. Only a lab test using the standard method is reliable.
Is tea with lemon dangerous in a decorated cup?
Acid draws metal out of glaze, which is why standards test with 4% acetic acid. A cup from a known factory that carries a standard number is no concern for a quick drink. The danger is leaving an acidic drink for hours in a craft-made vessel or a cup with worn decoration.
-
Sun K., Zhang N., Zhao J. et al. Determination of 24 elements migration in Yixing clay pottery by ICP-MS // 卫生研究. 2024. doi:10.19813/j.cnki.weishengyanjiu.2024.02.018. Abstract only. ↩
-
Xu Weixuan, Chen Wenbin. Determination of barium in food-contact zisha ware by energy-dispersive X-ray fluorescence spectrometry // 光谱学与光谱分析. 2023. Vol. 43(2). P. 475–483. Text. ↩
-
U.S. EPA. Lead Test Kits. epa.gov/lead/testkits. ↩