Do glass bottles have more microplastics than plastic ones?

In one French study, some drinks in glass bottles did carry more microplastics than the same kinds of drinks in plastic. The extra particles matched the paint on the metal caps, which points to the closure as the source. Water was low in both kinds of bottle, and the study didn’t look at health. Chaïb et al., 2025

What did the study find?

Researchers at ANSES, France’s food safety agency, bought water, cola, iced tea, lemonade, beer and wine from French shops in June 2023. The drinks came in glass bottles, plastic bottles, cans, cartons and bag-in-box packs. The team tested six containers of each product, all from one production batch. They counted particles under a microscope, identified a selection of them by infrared spectroscopy, and extended those results to the rest by shape and color. Chaïb et al., methods

That method has a floor. It counted particles from 30 to 500 micrometers across, and 500 micrometers is half a millimeter. Particles smaller than 30 micrometers were too hard to identify with the instrument, so they aren’t in these numbers.

For every drink except wine, glass bottles had the most particles:

Average particles per liter, 30 to 500 micrometers across. Source: Chaïb et al., 2025, results sections 3.4 to 3.10.

A few details change the picture:

• Beer. The high count came from small glass bottles. Large glass bottles, at 32.8 particles per liter, matched cans.
• Water. Water was low in every container: 4.5 particles per liter in glass and 1.6 in plastic. The paper counts that gap as statistically significant, and ANSES calls water relatively low whatever the container. Water in cartons from two brands averaged about 3. ANSES
• Wine. Wine was the exception, and the authors point out that its glass bottles were sealed with corks rather than metal caps. Wine in cartons had the highest wine count, 30.0 per liter.

Why would a glass bottle have more plastic in it?

Because of the paint on the caps. Most particles in the glass-bottled drinks were the same color as the paint on the outside of the caps. Infrared analysis showed the paint was mainly polyester, like most of those particles. The authors describe the caps as coated with an alkyd or polyester-based paint, and the study counts both as microplastics. Chaïb et al., sections 3.11 and 4.2

Under magnification, new caps had scratches in their painted outer surface and flecks of paint stuck to their inner face, the side that meets the drink. The authors think the caps rub against each other while stored in bulk before use, and ANSES describes that friction as the probable cause of the scratches. ANSES

Then the team tested the idea. They cleaned glass bottles, filled them with filtered water in which no particles could be found, and sealed them with new yellow caps. The caps went on straight from the supply, after being blown clean with air, or after being blown and then rinsed with filtered water and alcohol, ten times for each treatment.

Untreated caps left an average of 287 yellow paint particles per liter. Air cut that to 106, and air plus rinsing to 87, about 70% lower. The rinse liquid itself held about 48 yellow particles per cap. Cleaning helped a lot, but the authors note it didn’t remove every particle. Chaïb et al., section 3.11 and Figure 4

Those 287 particles per liter came from caps alone, in a test built to show the route, so they aren’t a measure of what the shop-bought drinks contained. ANSES suggests that bottlers clean caps before use, store them so they don’t rub, or change the paint. ANSES

So are plastic bottles the better choice?

Not on this evidence alone. In the French study, soft drinks in plastic bottles did have far fewer particles of this size. But the extra particles in glass traced to a cap problem that ANSES says bottlers can easily address, water barely differed, and other research complicates the picture:

• Glass-bottled water has surprised researchers before. A 2018 German study found some glass-bottled mineral waters with many particles, averaging 50 per liter (range 0 to 253), against 14 in single-use plastic bottles. Only returnable plastic bottles, at 118, stood clearly above the lab’s blank, and the authors suggested lubricants on caps as a source in the glass-bottled water. Schymanski et al., 2018
• Methods change the counts. Another 2018 German study could see particles down to 1 micrometer. It found microplastics in water from every bottle type, averaging 2,649 per liter in single-use plastic and 6,292 in glass with wide swings between bottles, and label-ink pigments in reusable bottles that pointed to bottle washing. More than 90% of its particles were smaller than 5 micrometers, far below what the French study could count. Oßmann et al., 2018
• The wider literature leans the same way, with exceptions. The French authors cite several studies that found more particles in glass containers and one that found fewer.
• Plastic caps shed too, as the next section shows.

Where these studies looked for a cause, they pointed at what surrounds the glass, such as caps, labels and washing, more than at the glass itself.

As for health, ANSES says it can’t tell whether these levels pose a risk, because toxicological reference data don’t exist yet. Our guide to the health evidence explains what the research can and can’t show, and our bottled water explainer covers the much larger counts from newer imaging methods.

Do plastic bottle caps shed microplastics too?

Yes. Lab studies find that opening and closing plastic caps wears particles off caps and bottle necks.

• Repeated opening. In a 2019 study, opening and closing PET water bottles with polyethylene caps 1, 10 and 100 times increased the particles on the neck and cap surfaces, most of all on caps after 100 openings, and brands differed. Squeezing the bottles didn’t raise the count in the water. Winkler et al., 2019
• Twisting caps open. A 2020 study twisted open about 40 washed, empty plastic bottles and detected particles, which the authors said could come from the tamper ring breaking or rubbing on the cap. Sobhani et al., 2020
• Glass bodies with plastic caps. A 2021 study opened and closed water bottles and baby bottles 100 times, all with plastic caps, then rinsed the necks and caps with ethanol to collect what came off. Glass-bodied water bottles released significantly fewer particles than plastic-bodied ones, and only 7.5% to 42.1% of the released particles were identified as plastic. Song et al., 2021
• Reusable bottles. A 2025 evidence map counted a study of reusable plastic bottles, where polypropylene particles from the screw caps rose with repeated opening, among its seven most reliable studies. Zimmermann et al., 2025

So the French finding adds a route rather than overturning anything: paint scratched off before the cap goes on, on top of the wear that comes with opening.

What does this mean for a reusable bottle?

Look past the body to the cap. Researchers who mapped this evidence in 2025 assumed that a glass bottle or jar likely has plastic in its closure, either a plastic cap or a plastic-coated metal one with a gasket. Zimmermann et al., methods Steel bottles vary too: Klean Kanteen, for example, makes many of its caps from polypropylene and also sells caps with an all-steel interior for its Classic, Wide and TKWide bottles, including the TKWide on our swap list. Klean Kanteen FAQ

The parts that touch your drink are the inside of the bottle, the underside of the cap, the threads and the seal. Where the threads are plastic, the studies above suggest that’s where wear happens. A steel-interior cap puts steel and a silicone ring on the drink side instead, so the seal is still a polymer part. No study we found has measured particles from caps like that, so we can’t promise they shed less; they do keep plastic threads and paint away from the drink.

What we don’t know yet

• How common these caps are. This was one set of drinks bought in France in 2023, with one batch per product. The paper names no brands, and we don’t know how many bottlers, in the US or elsewhere, use painted caps or already clean them.
• Anything smaller than 30 micrometers. The two German studies that could see smaller particles found most of theirs below 20 micrometers, so the French counts cover only part of the size range.
• The liners. The paper reports analyzing the caps’ outer paint, not their liners. Glass-bottled drinks also held more polyolefin particles, the plastic family that includes polyethylene, at up to 24.7 per liter, and the authors didn’t trace them.
• Plastic bottles and cans. ANSES says the origin of their particles remains to be explored.
• Additives. The authors note that paint can carry additives as well as polymers. They didn’t measure any.
• Health. There are no toxicological reference data for these particles yet. The WHO’s 2019 review judged that particles larger than 150 micrometers are unlikely to be absorbed and that uptake of smaller ones is expected to be limited, with very little data on the smallest sizes. WHO, 2019

What we’d do

1. Don’t switch containers over this headline. Water was low in glass and plastic alike. For soda and beer, the fix sits with bottlers. The study’s cleaning step happened before capping, so wiping a sealed cap at home can’t undo it. If you want to act on this one study for soft drinks, its plastic bottles and cans had fewer particles in this size range, though cans typically carry a plastic-like inner coating of their own. Zimmermann et al., methods
2. Fill your own bottle. That takes factory caps out of the picture. Tap water has its own particles, and our filtration guide covers what filters can and can’t do about them.
3. Keep the bottle you have if it works. Look at the cap from the inside and ask what the drink touches. Replace a cracked cap, and replace a seal that has stretched or torn.
4. If you want plastic and paint off the drink side, choose a steel or glass bottle whose closure is steel or glass on the inside, with a silicone or rubber seal. If your bottle’s maker sells such a cap, you may only need the cap.
5. Clean the cap and its seal the way the maker says, and check the seal when you do.

Our pick

Our criteria: the drink touches steel or glass plus a single seal, nothing on the drink side is painted or coated, and the maker publishes the materials and sells replacement seals. None of these has been tested for particle release, by us or by anyone we found. The first card is our main pick; the other two cost less and give something up.

Affiliate links may earn us a commission at no extra cost to you. How this works.

Klean Kanteen 27oz Reflect with Bamboo Cap

• Why it fits: It’s the closest match we found to the problem in this study: no paint anywhere, and a cap that is steel on the drink side.
• What touches your drink: The maker lists three materials. The 27-ounce bottle is 18/8 stainless steel, 90% recycled, with an electropolished interior. The cap is one piece of stainless steel on the inside with a bamboo top, and it seals with a food-grade silicone ring. Klean Kanteen says there is no paint, ink, coating or plastic. Klean Kanteen
• Evidence: Manufacturer-declared. No certification or independent test covers these claims.
• Cons: It’s single-wall, so it won’t keep drinks hot or cold, and the maker advises against hot and carbonated drinks in it. The bamboo cap is hand-wash only, shouldn’t be soaked and needs occasional oiling. The opening is a narrow 1.75 inches. It’s made in China, and the lifetime warranty covers manufacturing defects but excludes normal wear and tear. Replacement silicone rings cost $7.95 for two.
• Price: $32.95 at Klean Kanteen (checked Sept 2026).
• Skip it if: you want hot drinks or insulation, or your current bottle already does the job.

Klean Kanteen Classic Steel Loop Cap

• Why it fits: If you already own a Klean Kanteen Classic, this cap gives it a steel drink side for less than half the price of a new bottle, and unlike the bamboo cap it can go in the dishwasher.
• What touches your drink: 18/8 stainless steel, threads included, according to the maker. Klean Kanteen Its o-ring is food-grade silicone, per a Klean Kanteen staff answer on the cap maintenance kit page, not the cap’s own spec sheet.
• Evidence: Manufacturer-declared.
• What it gives up: It changes only the cap. The bottle’s body and outer finish stay as they are.
• Cons: It fits only Classic narrow-mouth bottles with internal threads, and it isn’t meant for hot drinks.
• Price: $14.95 at Klean Kanteen (checked Sept 2026).
• Skip it if: your bottle isn’t a Klean Kanteen Classic, or its current cap is fine.

WECK 766 Juice Jar, 1 liter

• Why it fits: For keeping water or juice in the fridge, the drink touches only glass and a rubber ring.
• What touches your drink: A glass jar and a glass lid, sealed by a natural rubber ring and held by metal clamps on the outside, according to WECK’s US importer. The jars are made in Germany. WECK
• Evidence: Manufacturer-declared.
• What it gives up: Portability. It’s a storage jar, and WECK says only water-bath canning makes the seal airtight, so don’t count on it in a bag.
• Cons: Rings and clamps are hand-wash only, and the rings stretch with oils and can dry-rot from dishwasher detergent. Glass can crack with sudden temperature changes; WECK advises keeping the difference under 122°F. Juice jars aren’t on WECK’s freezer-safe list, and shipping from its site is extra and limited to the 48 contiguous states.
• Price: $16.65 for three jars with lids, rings and clamps, plus shipping; spare rings are $2.10 to $3.00 for six (checked Sept 2026).
• Skip it if: you need something for a bag, or you have a latex allergy. The rings are natural rubber, and allergists advise people with severe latex reactions to avoid latex.

Resources to go further

• The French drinks study (Chaïb et al., 2025). Free to read, also on HAL. Section 3.11 and Figure 4 cover the caps.
• ANSES news release (June 2025). A short, plain summary from the agency whose lab ran the study.
• FCMiNo dashboard (Food Packaging Forum). Free, searchable data from the 2025 evidence map of particles released by food packaging, bottle openings included.
• Microplastics generated when opening plastic packaging (Sobhani et al., 2020). Open access, with microscope images of what twisting and tearing packaging produces.
• WHO key messages on microplastics in drinking-water (2019). What is known about absorption, and how little data exists for the smallest particles.
• Susan Freinkel, Plastic: A Toxic Love Story (Houghton Mifflin Harcourt, 2011). A readable history of plastic told through eight everyday objects, one of them the soda bottle.
• Matt Simon, A Poison Like No Other: How Microplastics Corrupted Our Planet and Our Bodies (Island Press, 2022). A science journalist’s reported tour of microplastics research. Its tone is more urgent than ours.