Water filters and microplastics: what actually works

Walk down the water-filter aisle and one word keeps repeating on every box: reduces. Reduces chlorine, reduces lead, reduces microplastics. What almost none of them tell you is which microplastics, at which size, and by how much.

That vagueness matters, because “microplastic” isn't one thing. It's a size range that runs from visible fragments down to particles smaller than a virus. A filter that catches a big polyester fiber won't necessarily stop a tiny nanoplastic sphere. So here’s an honest ranking of the main filter types in home kitchens, from best-supported to least, plus one kettle trick worth knowing.

1. Reverse osmosis: the strongest physical argument

Reverse osmosis (RO) units push water through a membrane so tight that even dissolved salts can't get through. That's a very small pore — small enough that any intact plastic particle, from a microplastic down to most nanoplastics, physically cannot fit.

In lab tests, tight membranes have removed more than 99% of nanoplastics by simple size exclusion ^[7]. In peer-reviewed testing of consumer point-of-use devices, removal of larger microplastic fragments landed between 78% and 100% across most tested technologies, with RO among the top performers ^[4].

Two caveats worth knowing. A torn or aged membrane fails to a much bigger effective pore, so cartridges have to be changed on schedule. And RO systems waste some water and typically need a dedicated faucet or under-sink install — they're the highest-effort option on this list.

2. Ceramic gravity filters: the underrated old technology

Ceramic filters work the way a very fine sieve does — a rigid porous block that physically blocks anything bigger than its pore size. They’re slow, they clog, and they need occasional scrubbing. They also, in a 2024 field study comparing household waters, produced the fewest microplastic particles of any point-of-use category tested — a little less than half a particle per liter, compared to about 1.1 for straight tap and 1.4 for multi-stage electric purifiers ^[5].

That’s one modest field study, not a definitive verdict. But the underlying idea is sound: when the barrier is a physical pore small enough to block the particle, the material almost doesn’t matter. Countertop gravity units using ceramic elements are a reasonable middle path between a pitcher and a full RO install.

3. Boiling (in a non-plastic kettle, in hard water)

This one surprised researchers too. A 2024 study in Environmental Science & Technology Letters found that boiling calcium-rich tap water for about five minutes causes microplastics to get trapped inside limescale as it forms, removing on the order of 80 to 90% of particles in a wide size range ^[2]. The authors defended the finding against methodological pushback later that same year ^[3].

The catch: it only works well in hard water. In soft water, removal drops to about 25% ^[2]. So this technique is regional — great advice in most of the American Midwest, much weaker in Seattle or Boston.

There’s a second catch. A 2025 study of polypropylene kettles found the plastic kettle itself sheds nanoplastics while boiling, worst in the first several uses and dropping off sharply after about 150 boils ^[8]. If you're going to use the boil trick, use a stainless steel or glass kettle, and pour through a cloth or fine strainer to catch the limescale. If you want to read more, check out our companion pieces: Your kettle, your coffee, and microplastics and Coffee and microplastics: Where they come from and how to reduce exposure.

4. Activated carbon pitchers: better for taste than for plastics

Most pitcher filters, refrigerator filters, and basic under-sink cartridges use activated carbon. Carbon is genuinely excellent at pulling out chlorine, taste-and-odor compounds, and many organic chemicals. It is not, on its own, a size-exclusion filter for particles.

Lab work shows that under typical tap-water chemistry, small negatively-charged nanoplastics can slip right past activated carbon without sticking ^[6]. Reviews of carbon-based materials for microplastic removal reach the same conclusion: plain granular carbon is inconsistent, and the impressive removal numbers you sometimes see usually come from specially modified carbon that isn’t in your grocery-store pitcher ^[11].

The starkest example: in the same peer-reviewed device test cited above, one pitcher using carbon plus ion-exchange resin put more microplastic particles into its output water than were in the input — it was net adding particles, not removing them ^[4]. Independent consumer testing by ConsumerLab in 2023 found large brand-to-brand differences among pitchers, and again noted at least one pitcher adding microplastics rather than removing them (the manufacturer said the filter hadn't been rinsed as instructed — take that as you will) ^[10].

A carbon pitcher is fine for taste. Don’t count on it as a microplastic barrier.

5. Refrigerator and faucet filters: mostly the same story

Fridge dispensers and screw-on faucet filters are almost all built around the same activated-carbon technology as pitchers. Some are certified under NSF/ANSI Standard 401, the “emerging contaminants” standard that includes a microplastic-reduction claim. That certification requires a device to reduce at least 85% of a specific half-micron test particle in a lab ^[1].

That’s a real bar, and worth respecting. But there’s no standard for particles smaller than half a micron, so nanoplastics aren’t tested at all. And a 2024 study using advanced microscopy on bottled water found something like a hundred to a thousand times more nanoplastic particles per liter than earlier microplastic-only counts had suggested — mostly in a size range no consumer filter has been certified against ^[12].

Translation: if a fridge or faucet filter carries the 401 mark, it's been shown to knock down a specific lab bead. That's better than nothing. It's not the same as a proven nanoplastic barrier.

6. Unfiltered tap: not zero, but not the worst

If none of this is an option, unfiltered municipal tap water is still generally better than single-use bottled water. A 2025 pilot study of a full-scale drinking water treatment plant confirmed that conventional treatment progressively reduces microplastics — not to zero, but meaningfully ^[9]. And repeatedly cracking open plastic bottles adds particles of its own. The best move you can make with tap water is to store it in glass or stainless steel, not to abandon it for bottled.

Notes on the science

A few honest limits worth naming. The NSF/ANSI 401 certification tests uniform lab-made polystyrene beads at 0.5 to 1 micrometer — real environmental microplastics are weathered, oddly shaped, coated in biofilm, and range far smaller. Between-lab agreement on microplastic counts is still poor: a 2024 international ring trial involving 84 labs reported between-lab variability of 45 to 129% on identical reference samples ^[13-14]. Two studies reporting “80% removal” and “99% removal” may not be measuring the same thing. Read percentages as directional, not precise.

Where Winnow fits in

Cutting exposure at the tap is a real, useful lever, and picking a better filter is one of the highest-yield changes you can make in a kitchen. Winnow is a probiotic consortium designed to work within the gut lumen. Its strains have been shown in laboratory testing to bind micro- and nanoplastics inside the gut. Filtering what enters the glass and giving whatever slips through something to bind to inside the gut are complementary steps, not competing ones.