Does Reverse Osmosis Remove PFAS?


Yes, reverse osmosis removes over 90% of PFAS, including the short-chain forever chemicals many carbon filters miss. See how well RO works, the NSF certification to look for, and its limits.


Updated: September 14, 2026
 / 
Jeremiah Zac
 / 

Share:


The short answer: Yes. Reverse osmosis (RO) removes PFAS well. The EPA reports that RO and nanofiltration membranes are "typically more than 90 percent effective at removing a wide range of PFAS, including shorter chain PFAS." (EPA) In peer-reviewed home testing, some units drove every PFAS compound below detection. (Patterson et al. 2019) PFAS are a group of synthetic "forever chemicals" that resist breaking down. RO is one of the strongest options for the short-chain versions that older filters miss. Two caveats matter: performance depends on an intact, well-maintained membrane, and certifications don't yet confirm treatment to the EPA's newest limits.

If you want the wider strategy for cutting these chemicals out of your home, start with our guide on how to avoid forever chemicals.

How Well RO Removes PFAS

RO forces water through a semi-permeable membrane with pores small enough to block most dissolved contaminants. The EPA states that RO and nanofiltration are "typically more than 90 percent effective at removing a wide range of PFAS, including shorter chain PFAS." (EPA) That short-chain point is the key advantage. Short-chain PFAS are smaller molecules that many carbon filters let slip through.

Independent testing backs the percentage up. A peer-reviewed study of point-of-use RO units found that two of three systems reduced every PFAS compound below the detection limit. (Patterson et al. 2019) The third unit still performed strongly: it removed 94.3 to 98.5% of PFOS, 97.3% of PFOA, 98.7% of PFHxS, and 76.7% of PFNA. (Patterson et al. 2019)

Bar chart: PFAS removal by a residential reverse osmosis unit. PFHxS 98.7%, PFOS 98.5%, PFOA 97.3%, PFNA 76.7%.
Per-compound PFAS removal for one point-of-use RO unit. Two other units in the same study reduced all PFAS below detection; the lower PFNA result traced to a leaking membrane seal. Source: Patterson et al., 2019.

The lower PFNA figure is worth understanding, because it tells you what actually governs performance. The researchers traced the weaker numbers to a leaking membrane seal, not to molecular size. (Patterson et al. 2019) Breakthrough correlated with membrane integrity, not chain length. (Patterson et al. 2019)

That's the takeaway that changes how you buy and maintain a system. A properly sealed, intact membrane is what delivers the results. A degraded seal or a worn membrane is what lets PFAS through. The chemistry is on your side; the hardware condition is the variable you control.

What to Look For

Certification is the fastest way to separate a real PFAS-reducing system from marketing language. Look for NSF/ANSI 58, the standard for reverse osmosis systems, which carries a PFOA/PFOS reduction claim. (NSF) A related standard, NSF/ANSI 53, covers non-RO filters and also carries PFAS reduction claims. State health guidance points to both: choose a product certified to NSF/ANSI 53 (for filters) or NSF/ANSI 58 (for reverse osmosis). (MDH)

There's an important limit to what certification currently proves. The Minnesota Department of Health notes that current certifications "do not yet indicate that a filter will remove PFAS down to the levels EPA has now set." (MDH) In plain terms, a certified system is tested to reduce PFAS, but the test thresholds haven't caught up to the EPA's newest limits.

So treat certification as a strong floor, not a guarantee of a specific outcome part-per-trillion. If your water is a known concern, test your treated water rather than assuming a number from the label. For product research, our roundups of the best reverse osmosis system for home and the best under-sink water filter list certified options across budgets. WWR earns a commission on some links; our recommendations stand independent of it.

Limitations

RO is effective, but it has real trade-offs you should weigh before buying.

It wastes water. In RO treatment, roughly 20% of the feedwater becomes concentrated waste. (EPA) That reject stream carries the rejected contaminants and goes down the drain. For homes on a well or watching usage, that's a running cost.

It needs maintenance. Effectiveness degrades if the unit isn't maintained. (MDH) This connects directly to the study finding above: an intact system performs, a neglected one leaks. Replace membranes and pre-filters on schedule, and check seals.

It removes dissolved minerals too. RO strips out beneficial minerals along with contaminants. This is often overstated as a health problem, and it's easy to address. Many systems add a remineralization stage, and a balanced diet covers the difference for most people. For most homes it's a manageable trade-off rather than a dealbreaker.

Why PFAS Matter

PFAS are worth treating because of what research links them to, stated carefully. The Agency for Toxic Substances and Disease Registry (ATSDR) says studies "suggest associations" between higher PFAS exposure and several health effects. (ATSDR) That wording matters: these are associations, not proven cause.

The effects ATSDR describes include raised cholesterol, lower antibody response to vaccines, changes in liver enzymes, pregnancy-related hypertension and preeclampsia, small decreases in birth weight, and kidney and testicular cancer. (ATSDR)

The regulatory response reflects that concern. In April 2024 the EPA finalized the first national drinking-water limits for PFAS, called Maximum Contaminant Levels (MCLs), measured in parts per trillion (ppt). (EPA) They set 4 ppt each for PFOA and PFOS, 10 ppt each for PFHxS, PFNA, and HFPO-DA (GenX), plus a Hazard Index of 1 for mixtures. (EPA) Those numbers give you a reference point for what "low" means with these chemicals.

Other Options That Remove PFAS

RO isn't the only effective method, and the right choice depends on which PFAS you're targeting. The EPA names two other treatments that work. (EPA)

Granular activated carbon (GAC) is strong on long-chain PFAS like PFOA and PFOS, but weaker on short-chain compounds. (EPA) It's a proven, widely used option, and often cheaper to run than RO. If carbon fits your situation, see our guide to the best activated carbon water filter.

Anion exchange is also effective at PFAS removal, per the EPA. (EPA) It works by swapping charged PFAS ions out of the water on a resin.

Here's how the three compare, based on EPA guidance (EPA):

Method Strongest at Weaker at Notes
Reverse osmosis Broad PFAS, including short-chain No major gap noted Wastes about 20% of feedwater; removes minerals
Granular activated carbon (GAC) Long-chain (PFOA, PFOS) Short-chain PFAS Often cheaper to run
Anion exchange High-capacity PFAS removal No major gap noted Typically higher cost than GAC

RO's edge over both is short-chain removal. (EPA) If your water shows a mix of PFAS or you want the broadest coverage in one system, that's the case for choosing RO.

2026 EPA Rule Changes

The federal picture shifted in 2026, so any advice keyed to the 2024 limits needs an update. On May 18, 2026, the EPA proposed to keep the 4 ppt MCLs for PFOA and PFOS but extend the compliance deadline to 2031. (EPA) In the same move, the agency proposed to rescind the limits for PFHxS, PFNA, GenX, and the Hazard Index. (EPA)

The EPA has announced it will keep the PFOA and PFOS levels while it reconsiders the rest. (EPA) As of September 2026, these are proposals in rulemaking, not final rules. (EPA)

Here's what that means for you at home. Utilities have more time and, for some compounds, potentially no federal limit. Point-of-use treatment puts that decision back in your hands regardless of where the rules land. An RO system that reduces a wide range of PFAS doesn't depend on which limits survive rulemaking.

Who Should Use RO

RO makes the most sense for a few specific situations.

  • Your water tests positive for PFAS, especially a mix. RO's broad, short-chain-capable removal covers more compounds in one system. (EPA)
  • You want treatment that doesn't hinge on shifting federal limits. Point-of-use RO reduces PFAS regardless of the rulemaking outcome.
  • You have a health-sensitive household. Given the associations ATSDR describes, some families choose the broadest available reduction. (ATSDR)

My recommendation: For most homes concerned about PFAS, an under-sink RO system certified to NSF/ANSI 58 with a PFAS reduction claim is the strongest single choice. (NSF) It handles both long- and short-chain PFAS, and the main trade-offs (water waste, maintenance, mineral removal) are manageable. If you're targeting only long-chain PFOA and PFOS and want lower running cost, a certified GAC filter is a reasonable alternative. (EPA) Whichever you pick, keep the seals and filters current, and test your treated water if PFAS is a known local problem.

This article is informational and not personalized medical advice. For decisions about your household's specific exposure or health, consult a qualified professional or your local health department. Drafts on WWR are AI-assisted and reviewed against our editorial standards.

Key Takeaways

  • Reverse osmosis removes more than 90% of a wide range of PFAS, including the short-chain types many filters miss. (EPA)
  • Verified home testing drove PFAS below detection in intact units; the weakest result came from a leaking membrane seal, not molecular size. (Patterson et al. 2019)
  • Buy a system certified to NSF/ANSI 58 with a PFAS reduction claim, and maintain the membrane and seals. (NSF)
  • Certifications don't yet confirm reduction to the EPA's newest limits, so test treated water if PFAS is a known concern. (MDH)
  • RO wastes about 20% of feedwater and removes minerals; both are manageable trade-offs. (EPA)
  • GAC (strong on long-chain PFAS) and anion exchange are effective alternatives; RO's edge is short-chain removal. (EPA)
  • As of September 2026, the EPA has proposed keeping the 4 ppt PFOA/PFOS limits (compliance extended to 2031) while rescinding several others; these are proposals, not final. (EPA)


Share this post!