How We Test Water Filters: Certifications, TDS and Lab Data

A water filter is only as good as the contaminants it is proven to reduce. So we start where most reviews skip: checking each product’s certification listings, contaminant by contaminant, in the official NSF, WQA, and IAPMO databases. Then we look at flow rate, capacity, waste water, and the real cost per gallon. When we test a filter ourselves, we publish the source water, equipment, lab, and sample dates.

Why certification checks come first

Certification is contaminant-specific. A filter certified under NSF/ANSI 42 for chlorine taste and odor is not automatically certified to reduce lead, and a filter certified for lead is not automatically certified for PFAS. Many product pages blur this with phrases like “tested to NSF standards” or “removes up to 99% of 2,000+ contaminants.” Our first job is to find out which claims are actually certified.

The research-based evaluation (every product)

Step 1: Pull the official certification listings

We search the model and its replacement cartridge in:

NSF Listings (NSF International)
WQA Gold Seal product directory (Water Quality Association, accredited by ANAB and SCC)
IAPMO R&T certification listings

We record the standard, each certified contaminant, the certified capacity (gallons), the certified flow rate, and the date we checked. We note whether the certification covers the whole system or only a component, and whether the listed model number matches the one being sold.

The standards we look for:

Standard	What it covers
NSF/ANSI 42	Aesthetic effects: chlorine, taste and odor, particulates
NSF/ANSI 53	Health effects: lead, cysts, certain VOCs, PFOA and PFOS, and a “total PFAS” claim
NSF/ANSI 58	Reverse osmosis systems: TDS plus optional claims such as arsenic, nitrate, PFAS
NSF/ANSI 55	UV systems (Class A for disinfection, Class B for supplemental treatment)
NSF/ANSI 62	Distillation systems
NSF/ANSI 401	Selected emerging contaminants such as certain pharmaceuticals and pesticides
NSF/ANSI 372	Low-lead materials in the product itself (not contaminant reduction)
NSF/ANSI 177	Shower filters (free chlorine reduction)

PFAS note. NSF’s older protocol P473 covered PFOA and PFOS reduction. Those requirements were folded into NSF/ANSI 53 and 58, and the standards now also include a “total PFAS” claim covering several compounds. If a product page still cites “P473,” we check whether the product holds a current 53 or 58 listing.

Step 2: Label every claim

In every review we sort claims into three groups:

Certified: appears on an NSF, WQA, or IAPMO listing for that exact model.
Tested by a named lab: the brand publishes a third-party lab report with the lab name, method, and date. We link to it and read the details, including the challenge concentration and number of gallons tested.
Manufacturer claim: no listing and no report we could find. We say so plainly.

“Tested to NSF standards” means a lab followed the test method. It does not mean the product is certified, and there is no ongoing audit. We explain this whenever a brand uses the phrase.

Step 3: Capacity and cost per gallon

We use the certified capacity where available, not a marketing figure, and calculate:

Cost per gallon = replacement filter price / certified capacity in gallons

Worked example. A pitcher cartridge costs $15 and is certified for 40 gallons (151 L). $15 / 40 = $0.375 per gallon. A household drinking 1 gallon a day would replace it about every 40 days and spend about 365 / 40 x $15 = $137 per year.

For multi-stage systems we add every stage at its own replacement interval, then show a five-year total including purchase price.

Step 4: Flow rate, GPD, and waste water

Pressurized filters (faucet, under-sink, whole-house): certified flow rate in gallons per minute (gpm).
Reverse osmosis: rated production in gallons per day (gpd), plus the pure-to-drain ratio.
Gravity and pitcher filters: time to filter one fill.

Reverse osmosis production depends on water pressure, temperature, and TDS. Ratings are usually given at specific test conditions, so we note them. Where a brand states a waste ratio such as 1:1 or 2:1, we flag whether that number comes from a certification listing or marketing.

Step 5: Maintenance, materials, and warranty

We note how cartridges are changed, whether the system uses proprietary or standard-size filters, NSF/ANSI 372 low-lead material certification, filter-change indicators, tank sanitizing needs for RO, and warranty length.

Step 6: Score and rank

Factor	Weight	What we look at
Certified contaminant reduction	45%	Certified claims matched to the use case (lead, PFAS, well water, etc.)
Cost of ownership	20%	Cost per gallon and five-year total
Flow, capacity, and waste	10%	gpm or gpd, certified capacity, RO drain ratio
Transparency	10%	Published performance data sheet, clear claims, no misleading “tested to” language
Installation and maintenance	10%	Setup difficulty, filter change ease, standard vs proprietary cartridges
Warranty and support	5%	Warranty length, filter availability

What we do not do

We do not describe a filter as “certified” unless we found it in an official listing.
We do not call a product “lab tested” without a real lab report.
We do not treat a TDS meter as a safety test.
We do not claim filtered water prevents illness. If you are worried about a health risk from your water, contact your water utility, local health department, or doctor.

Sources

NSF, Forever Chemicals and the Advancement of Filtration Standards
Water Quality Association, Product Certification and Find Gold Seal Certified Products
IAPMO R&T, Water Systems Certification
US EPA, Consumer Confidence Report Rule Revisions
Water Filter Guru, How We Test Water Filters (benchmark for lab before-and-after testing)