FilterProof

Guides

What each NSF standard actually covers

Seven standards govern home water treatment. Each covers a different thing, and each has a way of being misread. Here they are, with the trap in each.

Published 12 September 2026 · Checked against the certifiers' own listings

“Certified to NSF/ANSI 53” tells you less than it appears to. The standard contains around seventy separate reduction claims, and a product is certified for the ones it earned — sometimes one of them. The standard number is a heading, not a summary.

Below is what each standard covers, what it means in ordinary language, and the specific way each one gets misread. Claim counts are taken from NSF's own certification database.

Diagram showing NSF, WQA and IAPMO R&T as three independent certification schemes, each queried separately and reported separately rather than merged into a single verdict.
Different certifiers, different standards, different claim lists — kept separate, not merged.

NSF/ANSI 42 — Aesthetic effects

7 separate reduction claims · Chlorine, chloramine, taste and odour, particulates, zinc, bacteriostatic effects

In plain language: Makes water taste and look better. Nothing in this standard is about health.

The trap: This is the standard behind most “NSF certified” badges on pitchers and faucet filters. On its own it carries no health-effects claim — nothing about lead, PFAS or anything else under NSF/ANSI 53.

NSF/ANSI 53 — Health effects

70 separate reduction claims · Lead, cysts, mercury, asbestos, arsenic V, VOCs, atrazine, TTHM, radon, microcystin, PFOA, PFOS and Total PFAS

In plain language: The health standard. Every claim is separate and separately certified.

The trap: “Certified to NSF/ANSI 53” means certified for at least one of seventy claims — not all of them. Always ask which claim.

NSF/ANSI 58 — Reverse osmosis systems

47 separate reduction claims · TDS, arsenic, nitrate/nitrite, hexavalent and trivalent chromium, fluoride, lead, barium, cadmium, copper, radium, selenium, cysts

In plain language: The standard for reverse osmosis systems, which push water through a membrane.

The trap: NSF's own claim list offers no PFAS claim under this standard — but WQA does certify PFAS under it, and Coway and Kinetico hold PFOA and PFOS claims there. Certifier schemes are not identical, so “not certified by NSF under 58” does not mean “not certified”.

NSF/ANSI 401 — Emerging compounds

16 separate reduction claims · Microplastics, ibuprofen, naproxen, bisphenol A, estrone, DEET, phenytoin, metolachlor and others

In plain language: Trace pharmaceuticals, pesticides and industrial chemicals found at very low levels.

The trap: Often marketed as the premium tier. It is a genuine standard, but it does not include PFAS.

NSF/ANSI 44 — Cation exchange water softeners

4 separate reduction claims · Hardness, barium, radium 226/228, efficiency rating

In plain language: Water softeners. A scale problem, not a drinking water safety problem.

The trap: A softener is not a drinking water filter and is not a substitute for one.

NSF/ANSI 55 — Ultraviolet treatment

2 separate reduction claims · Class A disinfection, Class B supplemental bacteriostatic treatment

In plain language: Ultraviolet light to inactivate bacteria and viruses.

The trap: Class B is supplemental only and is not intended for water that is microbiologically unsafe. Class A is the one that disinfects.

NSF/ANSI 177 — Shower filters

1 separate reduction claims · Free available chlorine reduction

In plain language: Shower filtration, chlorine only.

The trap: A single claim exists under this standard. Everything else a shower filter promises is uncertified.

NSF/ANSI 372 — the one that is not about water

Not in the list above because it is not a water treatment standard at all. NSF/ANSI 372 certifies that a product's wetted materials are lead-free — that the housing and fittings will not add lead to your water. It is a materials standard borrowed from plumbing.

It is also routinely presented in a way that reads as lead removal. It is not. A product can be certified to NSF/ANSI 372 and have no certified lead reduction claim whatsoever.

Who is allowed to certify

NSF writes most of these standards, which is why its name is on them. But NSF is not the only body that can certify to them, and a product certified elsewhere is not certified less.

CertifierStanding
NSF
NSF International
Writes most of the standards and certifies to them. Public listing at info.nsf.org.
WQA
Water Quality Association Gold Seal
ANSI-accredited. Certifies to the same NSF/ANSI standards. Public listing at find.wqa.org.
IAPMO R&T
IAPMO Research and Testing
ANSI-accredited. Certifies to the same NSF/ANSI standards, plus its own dedicated PFOA/PFOS standard, NSF P473-2016. Public listing at pld.iapmo.org — searchable by listee, product description or standard, and each file's own listing page names its models and their reduction claims.
UL / CSA
UL Solutions, CSA Group
Also accredited to certify water treatment units.

This matters when you are checking a product. A brand absent from NSF's listing may be fully certified by WQA or IAPMO. Three searches, not one, before you conclude anything.

The gap worth knowing about — and the correction

In NSF's own claim list, PFAS reduction claims appear only under NSF/ANSI 53. There is no PFAS claim under NSF/ANSI 58, its reverse osmosis standard; we read the full list, and the only fluorine-related entry under 58 is fluoride, an unrelated compound.

It is tempting to conclude from this that PFAS cannot be certified under a reverse osmosis standard at all. That conclusion is wrong, and an earlier version of this page drew it. WQA does certify PFAS under NSF/ANSI 58 — Coway, Kinetico and Purunity each hold PFOA and PFOS claims there.

The real lesson is narrower and more useful: certifier schemes are not identical, so a single-database check produces false negatives. "Not in NSF's listing" means exactly that, and nothing more. In practice reverse osmosis performs well against PFAS, and "not certified for PFAS" and "does not reduce PFAS" were never the same statement anyway.