Best whole house water filter for PFAS: which units carry an NSF/ANSI 53 PFOA/PFOS claim, by published spec

If nobody in the house drinks or cooks from anything but the kitchen tap, you have permission to leave: an under-sink RO system certified to NSF/ANSI 58 is the route EPA’s guidance actually describes, and it costs less than one year of cartridges for the units below.

How was this shortlist built?

Every figure on this page is the manufacturer’s published specification, read from the sell sheet or product page named beside it on 2026-09-14, or EPA’s rule text. Nothing was bench-tested.

The candidates were compared on the specs that decide a PFAS purchase: who certified the PFAS claim and to which standard, the exact wording of the claim, gallon capacity, rated flow in gallons per minute (gpm), pressure range, and whether the unit is a single tank or a cartridge housing.

“Certified” on this page means the maker’s own page names a certification body and a standard number; no certificate was pulled from the NSF or WQA databases this session. The NSF/ANSI standards guide explains what 42, 53, 58 and P473 each cover.

Two makers were dropped from the picks. Express Water publishes no certified or lab-tested PFAS claim for the WH300SCKS that could be found, and Waterdrop’s WHF21-FG is an iron, manganese and chlorine unit certified to NSF/ANSI 372 for lead-free materials only.

Why is the kitchen tap the certified route?

EPA’s page on filters certified to reduce PFAS names two standards, NSF/ANSI 53 and NSF/ANSI 58, and addresses point-of-use systems only. It also says that, as of April 2024, the existing certification standards do not yet show a filter reaching the 4.0 ppt level EPA set.

That is the honest flaw in every unit below. A whole-house tank filters every fixture, including the toilets, and is sized and priced for that. A pitcher certified to NSF/ANSI 53 or an under-sink RO filters the two gallons a day the household drinks, and the certified list for those is long. The PFAS guide carries that list and how to read a PFAS lab result.

Buy whole-house only when the reason is the whole house: a lab result above 4.0 ppt and a household that will not drink from one tap.

How do the units compare on paper?

Manufacturer figures. “Not published” means the page read this session prints no value, not that none exists. No unit’s price is copied from a listing.

A.O. Smith HomeShield AOS-HS-1200: buy it or skip it?

Buy it if the deciding line is an NSF/ANSI 53 certification for PFAS on the whole-house unit itself. A.O. Smith’s sell sheet is the only maker document found that prints that: below 4 ng/L for 500,000 gallons at 5.7 gpm, with PFNA, PFOA, PFOS, PFHpA and PFHxS footnoted as the compounds covered.

The sheet gives a 25 to 100 psi operating range and a 15.5 by 59.5 inch single tank with no drain line, so nothing goes to waste and no backwash is scheduled. A.O. Smith also prints a 500,000 gallon chlorine and chloramine capacity at 3 ppm, labelled as internal validation rather than a WQA-certified figure.

Skip it if the house runs sustained flow above 5.7 gpm: the sell sheet’s 11.2 gpm at 15 psid is marked not WQA-tested, and A.O. Smith says capacity may fall when the unit runs there. The whole-house flow rate calculator gives the household’s peak number. Skip it if a price is needed before calling a dealer: none is published, and no Amazon listing for the model was found.

Aquasana OptimH2O: buy it or skip it?

Buy it if lead and PFOA/PFOS are both on the lab report. Aquasana’s press release says IAPMO tested and certified the OptimH2O to NSF/ANSI 53 for soluble and particulate lead and cysts and to NSF Standard P473 for PFOA/PFOS, and Aquasana’s certified-products page lists up to 98% of PFOA/PFOS, up to 99% of lead and up to 90% of chlorine and chloramines.

NSF P473 is the earlier NSF protocol for PFOA and PFOS reduction; it predates the PFAS reduction claim inside NSF/ANSI 53 and tests to the older EPA health advisory level, not the 4.0 ppt MCL. Aquasana’s replacement page puts the main filter at 100,000 gallons or 12 months.

Skip it if a rated gpm or pressure drop has to be on paper: no Aquasana page printed either. Skip it if the purchase has to go through Amazon: both listings checked were marked unavailable and are not linked here. It is not the Aquasana Rhino EQ-1000 on the city-water whole-house page, a chlorine unit with no PFAS claim.

iSpring WGB32B-PF: buy it or skip it?

Buy it if the budget is $699.99 at iSpring’s list price and a lab-tested claim is acceptable in place of a certified one. iSpring’s page says the system is SGS-tested to remove up to 99% of PFOA and PFOS, with a capacity of 100,000 gallons or 12 months, in three 20 by 4.5 inch cartridges on 1 inch NPT ports.

SGS is a private testing laboratory, and iSpring’s page cites no NSF/ANSI standard number for the PFAS claim, so this unit is in the table as tested, not certified.

Skip it if only an NSF/ANSI 53 mark will do. Skip the plan of buying the standard WGB32B on Amazon and adding PFAS media later: Amazon carries the WGB32B, WGB32B-PB, WGB32B-KS and WGB32BM, none of them the -PF SKU, and the WGB32B parts page covers only the standard cartridges.

SpringWell Whole Home PFAS Filter: buy it or skip it?

Buy it only after SpringWell’s own page has been read: it returned an error to every fetch this session, so the figures here come from Applied Cartridge Systems (ACS), maker of the ATOMUS PF8 cartridge third-party sources report SpringWell packages. Skip it if “internally tested” is not enough, since the PFAS result is ACS’s own test, not a third-party certification, and skip it until SpringWell’s own capacity, gpm and price are confirmed.

ACS’s brochure says the PF8 was internally tested to the NSF/ANSI 53 protocol at 7 gpm across two filters, reaching non-detect, below 1 ng/L, for 8 PFAS compounds. It rates one cartridge at up to 5 gpm and 350,000 gallons at 100 ppt influent, or 150,000 gallons and more at 3,000 ppt. The NSF certifications ACS holds, NSF/ANSI 61 and 372, cover material safety and lead content, not PFAS.

What goes wrong?

The percentage is read as a guarantee. From 300 ppt of PFOA, 98% leaves 6 ppt, above EPA’s 4.0 ppt MCL, and only the A.O. Smith sheet and the ACS brochure state an outlet concentration at all.

Capacity is spent faster than the gallon figure suggests: a high lab result shortens the interval, as ACS’s own split between 100 ppt and 3,000 ppt influent shows.

The unit runs past its certified flow, since a shower and a washing machine together sit nearer A.O. Smith’s untested 11.2 gpm figure than its certified 5.7 gpm.

The wrong SKU arrives. The WGB32B on Amazon is not the WGB32B-PF, the Rhino EQ-1000 is not the OptimH2O, and a Waterdrop WHF21-FG is an iron and chlorine filter with no PFAS claim at all.

Frequently asked questions

Sources

1. EPA, PFAS National Primary Drinking Water Regulation (government or regulatory source)
2. EPA, Identifying Drinking Water Filters Certified to Reduce PFAS (government or regulatory source)
3. A.O. Smith, HomeShield AOS-HS-1200 sell sheet (PDF) (manufacturer or vendor document)
4. Aquasana, press release: whole house water filter certified to reduce lead and PFOA/PFOS (manufacturer or vendor document)
5. Aquasana, Certified Products listing (manufacturer or vendor document)
6. Aquasana, OptimH2O Whole House Filter Replacement (interval) (manufacturer or vendor document)
7. iSpring, WGB32B-PF PFAS whole house water filter system product page (manufacturer or vendor document)
8. Applied Cartridge Systems, ACS ATOMUS PF8 PFAS removal cartridge brochure (PDF) (manufacturer or vendor document)
9. Pennsylvania DEP, Point of Use Units NSF/ANSI 53 Certified for PFAS Reduction (PDF) (government or regulatory source)