Yes. A properly functioning reverse osmosis (RO) system removes PFOA and PFOS from drinking water with high efficiency, typically rejecting well over 90 percent of both compounds under normal operating conditions. The separation happens at the molecular scale: the membrane's polymer network excludes these relatively large, charged molecules while allowing water to pass. The word doing the heavy lifting is "properly" — rejection falls when the membrane degrades, when the system runs at low pressure, or when water chemistry shifts.

The Size-Exclusion and Charge Mechanism

Reverse osmosis works by forcing water through a semipermeable polymer membrane against its natural osmotic flow. The membrane is not a simple sieve. It is a dense, cross-linked polyamide layer, roughly 100 to 200 nanometers thick, that water molecules dissolve into and diffuse across under pressure. Anything that dissolves poorly in that polymer, or that is physically bulky or electrically repelled, stays behind in the concentrate stream.

PFOA (perfluorooctanoic acid) and PFOS (perfluorooctane sulfonic acid) are a good fit for what RO rejects. Each molecule carries a fully fluorinated eight-carbon tail and a charged acid head group. Molecular weights sit around 414 grams per mole for PFOA and 500 for PFOS — several times heavier than water. In water at neutral pH, both exist almost entirely as anions, giving them a negative charge the membrane also repels.

Size exclusion alone does not explain the high rejection. RO membranes are not tight enough to block every molecule in that weight range. Charge repulsion does substantial work: the membrane surface carries negative charge, so the anionic PFOA and PFOS are electrostatically pushed back. That is why rejection of these compounds tracks so closely with pH and ionic strength.

The plain-language version: the membrane filters by size and by electrical charge at once, and PFAS molecules lose on both counts.

Quantified Context: How Well RO Actually Performs

Reported rejection rates for PFOA and PFOS under controlled conditions cluster in the high 90s as a percentage. Studies of thin-film composite membranes consistently show rejection above 95 percent for both compounds when the membrane is intact, feed pressure is adequate, and pH is in the neutral-to-basic range. Some bench studies report rejection above 99 percent.

The number that matters for a household is different from the bench number. A single-pass under-sink RO unit is not a sealed laboratory cell. Real rejection depends on the ratio of product water to concentrate water, membrane condition, and temperature. Cold water lowers flux and can shift rejection. A unit producing very little wastewater per gallon of product forces more contaminants across the membrane per unit of water.

Concentration matters too. Higher feed concentrations of PFAS slightly reduce percent rejection on some membranes, so a system tested at trace levels may perform a little differently in heavily contaminated source water. Contamination in U.S. drinking water is typically measured in parts per trillion, not parts per million, which keeps the feed load low and rejection stable.

Timeline: Where RO Sits in PFAS Treatment

Stage What happens Practical implication
Contaminant enters feed water PFOA and PFOS arrive as dissolved anions Dissolved contaminants, not particles, so sediment filters miss them
Pre-filtration Sediment and carbon stages remove chlorine and larger matter Protects the RO membrane from chlorine degradation
RO membrane passage Water diffuses through; PFAS rejected by size and charge Main removal step, high-90s percent rejection
Concentrate rejection PFAS flushed to drain in wastewater stream Contaminants leave the system, not stored on the membrane
Post-treatment Polishing carbon improves taste Minor PFAS back-diffusion can occur here
Storage and dispensing Treated water waits in a tank Tank and lines can be a recontamination point if not maintained
Membrane aging Fouling, scaling, and chlorine attack degrade rejection Rejection is not permanent; replacement restores it

The counterintuitive entry is the post-treatment stage. The carbon polishing filter that improves taste sits downstream of the membrane, and small amounts of PFAS can migrate backward into stored water from that carbon or from the tank. The membrane does the real work; the carbon is a finishing step, not a safety barrier.

Edge Cases That Change the Answer

Rejection is not a fixed property of a device. It varies with operating conditions in ways that matter for anyone relying on RO for PFAS removal.

Chlorine exposure is the most common failure mode. Thin-film composite polyamide membranes degrade under continuous chlorine contact. Municipal water is chlorinated, so a system without adequate carbon pre-treatment will see its membrane lose both flux and rejection over time. A degraded membrane can still produce water that tastes fine while letting more PFAS through.

Low line pressure is another. RO needs pressure to overcome osmotic pressure and drive water across the membrane. Under-pressurized systems in older homes or gravity-fed setups run at reduced rejection. Temperature cuts both ways: cold feed water lowers production and can lower rejection, which is why winter performance sometimes differs from summer.

Feed water pH shifts the charge on both the membrane and the PFAS molecules. Higher pH increases the negative charge on PFOA and PFOS, generally improving electrostatic rejection. Very low pH can protonate the acids and reduce that charge advantage.

Mixtures and co-contaminants add complexity. High levels of dissolved solids, hardness, or organic matter foul the membrane and create pathways around it. A fouled membrane is not a uniformly performing membrane.

Common Misconceptions

The most widespread error is treating "RO" as a single specification. Membrane material, system design, pre-treatment, and maintenance all determine real-world rejection. Two units both labeled "reverse osmosis" can perform very differently.

A second misconception is that any carbon filter removes PFAS. Granular activated carbon has some affinity for longer-chain PFAS, but it is not a reliable primary barrier, and its capacity is finite and hard to judge. RO is the more defensible barrier for these compounds.

A third is that boiling water helps. Boiling concentrates PFAS because the compounds are not volatile; it makes the problem worse per volume. Distillation can remove them, but it is energy-intensive and slow at household scale.

A fourth: "the filter is certified, so I'm covered." Certification for PFAS reduction is a specific claim. A system certified for taste and chlorine is not thereby certified for PFOA and PFOS.

Watch for these red flags when evaluating a system: No third-party certification specifically listing PFOA and PFOS reduction. No carbon pre-filter ahead of the membrane in a chlorinated water supply. No published membrane replacement interval or pressure specification. Claims of "100 percent removal" of anything, which no membrane system honestly makes. A system sold without a way to monitor or replace the post-carbon stage.

RO Compared With Other PFAS Treatment Options

Method Typical PFAS reduction Household practicality Notes
Reverse osmosis High-90s percent under good conditions Point-of-use, under sink Requires pressure, pre-treatment, maintenance
Granular activated carbon Variable, partial for longer-chain PFAS Common, low cost Not a reliable sole barrier
Ion exchange resin High for many PFAS Point-of-use and whole-house Regeneration and disposal concerns
Distillation High Slow, energy-heavy Impractical for daily household volume
Nanofiltration Moderate to high Less common residentially Looser than RO; performance varies
Boiling None — concentrates PFAS N/A Counterproductive for PFAS

The comparison clarifies the practical hierarchy: RO and ion exchange are the strong household barriers, carbon is supporting cast, and boiling is the wrong tool entirely.

Practical Actions and Thresholds

The U.S. Environmental Protection Agency set legally enforceable maximum contaminant levels for PFOA and PFOS at 4 parts per trillion each in 2024, with a hazard-index approach for mixtures of several PFAS. Some states set their own, sometimes stricter, limits. The practical takeaway from reconciling these thresholds: any detectable PFAS at the parts-per-trillion level is treated as a health concern under current U.S. regulation, so a treatment goal of "as low as reasonably achievable" is the right frame, not "below some comfortable number."

Concrete steps: Check your utility's Consumer Confidence Report for PFAS results; if you use a private well, test for PFOA and PFOS through a certified lab. Choose an RO system with third-party certification specifically naming PFOA and PFOS reduction. Confirm the system includes adequate carbon pre-filtration if your water is chlorinated. Replace pre-filters and post-carbon on the manufacturer's schedule, and replace the membrane when production drops or rejection is suspect. Test treated water occasionally if source contamination is known, rather than assuming performance holds indefinitely.

The cost picture is modest at household scale. Under-sink RO systems commonly run from roughly USD 150 to USD 500 for the hardware, with annual filter and membrane costs in the tens of dollars. That is inexpensive insurance against a contaminant class measured in parts per trillion.

Limitations and What RO Cannot Do

RO removes PFOA and PFOS from the water you drink, but it does not solve contamination. The concentrate stream carries the rejected PFAS to the drain, and from there into wastewater systems that are not generally designed to destroy PFAS. Point-of-use treatment protects the household, not the watershed.

Whole-house RO is rare and expensive because it must treat every gallon used for laundry, irrigation, and bathing at high flow. Most PFAS exposure routes of concern are drinking and cooking water, which point-of-use treatment covers well.

Performance claims age. A certification reflects performance when tested, not a lifetime guarantee. Membranes foul, carbon exhausts, and seals fail. The honest framing is that RO is a highly effective barrier that requires maintenance to stay one.

Not all PFAS are identical either. Short-chain compounds behave differently from PFOA and PFOS, and rejection data for the long-chain compounds does not automatically transfer to every member of the class. For PFOA and PFOS specifically, the evidence for high RO rejection is strong. For the broader universe of PFAS, treat the answer as "likely effective but less certain" and verify against certification data.

One professional caveat worth stating plainly: percent rejection is a ratio, and ratios hide absolute numbers. A 99 percent rejection rate on heavily contaminated water can still leave more PFAS in the product than a 95 percent rate on lightly contaminated water. Source concentration and rejection rate have to be read together, which is why testing the treated water is more informative than trusting a specification sheet.