Do water filters remove PFAS?

What to look for on the box: legally required labels and the industry’s voluntary labels

On a water purifier’s box and manual, labels required by law sit side by side with labels based on an industry association’s voluntary standard. PFOS and PFOA belong to the second group.

• Under the Household Goods Quality Labeling Act (“家庭用品品質表示法”) and its labeling rules for miscellaneous manufactured goods (“雑貨工業品品質表示規程”), the “substances removed” that a water purifier may list are limited to 12 terms: free residual chlorine, turbidity, seven volatile organic compounds such as chloroform, CAT (a pesticide), 2-MIB (a musty odor) and soluble lead. PFOS and PFOA are not among them (checked on the Consumer Affairs Agency’s page in October 2026).
• The legally required label takes the form “rated volume (L), 80% removal, JIS S 3201 test result.” The rated volume (total filtered water) is the amount of water that can be filtered before the removal rate falls to 80%.
• PFOS and PFOA claims are optional labels based on JWPAS B, a voluntary standard of the Japan Water Purifier Association (JWPA). For example, Toray writes “PFOS and PFOA: rated volume 600 L (80% removal, JWPAS B.210 test result)” separately from the legally required section.
• The JWPA says it is “considering” adding a PFOS and PFOA test method to JIS (Japanese Industrial Standards). As of October 2026, we found no document confirming that a PFOS and PFOA test has been added to JIS.
• The JWPA publishes a “registry of household water purifiers that can remove PFOS and PFOA” (19 companies). Panasonic and BRITA are not on this list, but they state on their own pages that their products remove 80% or more under JWPAS.
• Products whose filtering function is an add-on, such as alkaline water ionizers, are outside this law’s water purifier labeling (as explained by the JWPA).

PFOS and PFOA claims by product

Only claims confirmed on the makers’ official pages and catalogues are listed. The rated volume is the amount of water treated before removal falls to 80% when water containing 50 ng/L of PFOS and PFOA combined is passed through (JWPAS B.210, the Japan Water Purifier Association standard).

How each type works

What does “80% removal” mean? (The JWPAS B.210 test)

Key points of the PFOS and PFOA test method published by the JWPA’s technical committee (October 11, 2023).

• It is a test in which up to 10 ng/L may remain in the filtered water at the end of the cartridge’s life (20% of 50 ng/L; Urayomi’s calculation). That is one fifth of Japan’s water quality standard (PFOS + PFOA 50 ng/L), but higher than the US EPA standard (4.0 ng/L each for PFOA and PFOS). For how the standards differ, see How was the “50 ng/L” standard set?.
• KITZ explains that 80% means “removes 80% or more even at the end of its life,” not that it always removes only 80%. In test examples published by the JWPA (faucet-mounted and faucet-integrated types, third-party testing laboratory), test water of 47–51 ng/L stayed below 1 ng/L up to the stated rated volume (removal rate of 97.9% or more). The JWPA notes that this does not guarantee that every water purifier performs the same.
• If your area’s tap water is X ng/L, then even if the removal rate falls to 80% at the end of the cartridge’s life as in the test, the filtered water would be about 0.2X ng/L as a rough guide (Urayomi’s calculation; removal can change when the concentration differs, so this is only a guide). Find your area’s values from How can I check PFAS in my area’s tap water?.
• Only two substances, PFOS and PFOA, are tested. This label does not tell you how well other PFAS such as PFHxS are removed.
• Sources disagree on when PFOS and PFOA were added to JWPAS: “revisions in August 2022 and May 2023” (Panasonic) versus “test method established in October 2023” (JWPA).

Rated volume and cartridge cost per liter

Of the 27 products in this table, 12 products have a PFOS and PFOA rated volume published by the manufacturer. Cleansui, LIXIL, BRITA, Duskin, SANEI and Waterstand write only “removes 80% or more,” so their volume is unknown.

• Numbers on the box such as “rated volume 900 L” are often values for chlorine or JIS substances and are not necessarily the same as the volume for PFOS and PFOA.
• Dividing the cartridge price by the PFOS and PFOA rated volume gave a range from about ¥0.34 (KITZ GOQURIA) to about ¥3.19 (OSG Twin e-S, assuming both cartridges are replaced) per liter (Urayomi’s calculation). We did not calculate it for Toray’s pitcher and faucet-mounted models because their cartridges have open prices.
• The replacement period is calculated by assuming how many liters are used per day. For example, Toray gives 6 months for the PT307SLV, based on 600 L at 3 L a day. If you use 6 L a day, you reach 600 L in about 100 days (Urayomi’s calculation).
• For the Panasonic TK-CJ24, the replacement guideline differs by substance: about 1 year for PFOS and PFOA, and about 6 months if the purpose is removing chloroform and similar substances. Zenken and Duskin give replacement guidelines based on the substance with the shortest purifying capacity.

How is this different from “NSF certification”?

NSF is a US certification body. Its PFOA and PFOS tests for water filters use much more concentrated water than Japan’s JWPAS.

• The test water concentrations differ by about 30 times. JWPAS is closer to the levels in Japanese tap water (a few to several tens of ng/L), while NSF assumes the high concentrations of contaminated areas. Even the same “80%” or “95%” is measured under different conditions, so neither can be called better.
• It has been reported that the 2025 edition of NSF/ANSI 53 updated the treated-water limits for PFOA and PFOS to match the US EPA standards, but we could not confirm the figures in the text of the standard.
• Even under the same “NSF/ANSI 53 certification,” whether PFOA and PFOS items are included differs by product, and so does the edition of the standard used for the evaluation. For Multipure, the replacement element CBXT (listed as CB-EXTRA) and the Aquaperform rental unit (CB11As), among others, have PFOA and PFOS items, but the Aquaversa (CB6) does not.
• Searching the NSF listings by company name (October 3, 2026), we did not find Panasonic, Takagi, Zenken, LIXIL, TOTO, KITZ, Meisui, OSG, Duskin, Nihon Trim, Waterstand, Toclas, SANEI or KVK. Mitsubishi Chemical Cleansui and BRITA SE are listed, but without PFOA or PFOS items. Because entries can be missed due to how a company name is written or due to OEM (manufacturing under another company’s name), we cannot go so far as to say they are “not certified.”
• Amway’s eSpring is listed under NSF/ANSI 53 for items including PFOA and PFOS, but we could not confirm how it corresponds to the model numbers sold in Japan.

How much did filters remove in studies?

Manufacturer tests are lab results under fixed conditions. There are also studies of water filters actually used in homes, and studies comparing products on the market.

• The US EPA explains that activated carbon adsorbs long-chain PFAS such as PFOS and PFOA well, but does not adsorb short-chain PFAS such as PFBS and PFBA very well. RO and nanofiltration typically remove 90% or more, including short-chain PFAS.
• The Japanese labels (JWPAS) are based on tests of PFOS and PFOA only. For most products, how well short-chain PFAS are removed has not been published.
• In a cross-sectional study in Massachusetts, US (Heckel et al. 2026), people in one area who drank filtered tap water or bottled water had lower PFAS levels in their blood, but in the other area there was no difference. It is not a study showing that water filters lower blood levels. For health effects, see What effects does PFAS have on the body?.

If you don’t replace the cartridge

A cartridge’s performance declines little by little from the first use. The stated rated volume is the volume “until removal falls to 80%,” and performance beyond that point is not tested.

• In the study of US homes (Herkert et al. 2020), 4 of 8 whole-house activated carbon systems had more PFAS in the water after filtering than before (the abstract does not give the cause).
• A review (MacKeown et al. 2024) also says that activated carbon that is not maintained can actually increase PFAS.
• The replacement period in months is calculated by assuming a fixed daily use. In households that use more, for example for cooking as well, the rated volume is reached sooner than the number of months suggests.
• Meisui gives the replacement time for the nomot in both volume and months: “3,000 L or within 12 months.” If you do not know how much you have used, replacing at whichever comes first, the months or the volume, is the use that matches the label’s assumptions.

Counterfeits and used cartridges

• On September 26, 2024, the Japan Water Purifier Association (JWPA) issued a “Declaration on Strengthening Measures Against Counterfeit Filter Cartridges” (“浄水カートリッジ模倣品対策強化宣言”). It says counterfeits have elaborate packaging and appearance that make them hard to tell apart, and that if they lack the original function and quality, they could even harm health.
• Takagi states clearly that it does not sell genuine cartridges on flea-market or auction sites or on major online shopping sites (except its official store). When a manufacturer names its official sales channels, buying there is the reliable choice.
• For used cartridges, BRITA Japan says that collection and recycling at the same level as in Germany is not possible in Japan, and advises disposing of them according to local rules. Check the waste sorting category with your city, ward, town or village. We did not check whether other manufacturers collect used cartridges.
• A rough estimate of the PFOS and PFOA that builds up in a cartridge: even if 4,000 L of 50 ng/L water passed through and all of it was adsorbed, it would be 200 μg, or 30 μg for 600 L (pitcher) (Urayomi’s calculation).

Related explainers

• Japan’s PFAS standard for tap water, “50 ng/L”: how was it set, and how does it differ from the U.S. and EU?
• How do you check PFAS in your local tap water? Reading the national tables and working out “what percent of the TDI”
• Does boiling tap water reduce PFAS? What public bodies and research say
• What about PFAS in mineral water and bottled water? Japan’s specification and the national survey