There is no single best household water filter. Each of the three main mechanisms — activated carbon, reverse osmosis, and gravity — is the right answer for a different reader's priorities. The frame that matters is: which contaminants do you actually care about, what is your budget over five years, what install can your kitchen accommodate, how much water waste are you willing to tolerate, and how do you feel about flat-tasting demineralized water.
This article is the technology-first version of the comparison. Once the mechanism is clear, the 10 products in our water catalog map cleanly to the use cases.
The three mechanisms — what each actually does
Activated carbon (granular, block, or catalytic) works by adsorption and, in the catalytic variant, surface chemistry. It captures chlorine, VOCs, taste-and-odor compounds, and, in NSF/ANSI 53 certified models, lead and a growing list of PFAS. NSF describes Standard 53 as the health-effect standard covering lead, VOCs, cysts, chromium, and now PFOA, PFOS, and four additional PFAS as a "Total PFAS" reduction claim. What carbon does not do well: dissolved minerals, total dissolved solids, fluoride (without an alumina or bone-char additive), nitrate, and sulfate. Standard pitcher carbon also tends to underperform versus carbon block on contact-time-dependent contaminants like chloramine.
Reverse osmosis (RO) uses a semi-permeable membrane to physically reject almost everything dissolved in the water — heavy metals, fluoride, nitrate, perchlorate, microplastics, and PFAS. NSF/ANSI 58 is the standard that certifies point-of-use RO systems for total dissolved solids and the listed contaminant claims. The trade is two-fold: minerals come out along with contaminants, leaving the water flat-tasting unless remineralized; and RO produces reject water — the brine the membrane rejects. The EPA WaterSense program notes typical POU RO systems send multiple gallons down the drain for every gallon delivered, and labels only systems that achieve a waste-to-product ratio of 2.3 or lower.
Gravity systems — Berkey-style — sit on the counter and pass water through dense ceramic and carbon "elements" by gravity alone. No electricity, no plumbing. Element pore sizes are large enough for water but small enough to physically catch sediment, bacteria, and cysts, and the carbon component handles chlorine, many VOCs, lead (per disclosed testing), and some PFAS depending on element specification. Mineral content is preserved more than RO but slightly reduced versus simple carbon.
The contaminant matrix
| Contaminant | Carbon (block / catalytic) | Reverse osmosis | Gravity | | --- | --- | --- | --- | | Chlorine | Yes — primary use case | Yes | Yes | | Chloramine | Catalytic carbon only | Yes | Element-dependent | | Lead | NSF 53 certified models | Yes | Per disclosed testing | | PFAS (PFOA / PFOS) | NSF 53 / P473 certified blocks | Yes | Per disclosed testing | | Fluoride | Only with alumina / bone-char additive | Yes | Optional fluoride element | | Nitrate | No | Yes | No | | Microplastics | Block partial | Yes | Ceramic element yes | | Bacteria / cysts | No (cysts only on cyst-rated blocks) | Yes | Yes — physical exclusion | | Minerals retained | Yes | No (unless remineralized) | Largely yes |
The certifications to ask for, in order: NSF/ANSI 42 covers aesthetic effects (chlorine taste, odor, particulates). Standard 53 covers health-effect contaminants. Standard 58 covers RO performance. Standard 401 covers emerging compounds like pharmaceuticals and pesticides. Protocol P473 is the targeted PFOA / PFOS test that became the public-facing PFAS marker before NSF 53 expanded its PFAS coverage.
The EPA's PFAS treatment guidance lists granular activated carbon, anion exchange, and reverse osmosis as the three effective technologies — which is exactly the spread our matrix above maps to.
Cost over five years
These are approximate ranges based on manufacturer-disclosed replacement schedules. Real-world filter life varies with source-water sediment load and household volume.
- Pitcher carbon (Brita Elite, ZeroWater, Epic Pure): $30–$55 unit + roughly $40–$80 per year in cartridges. Five-year envelope: about $230–$455.
- Premium pitcher carbon (Clearly Filtered): about $90 unit + $80–$120 per year. Five-year envelope: about $490–$690.
- Under-sink carbon (Hydroviv, Aquasana AQ-5300, Crystal Quest): $300–$700 unit + $120–$200 per year. Five-year envelope: about $900–$1,700.
- Countertop RO (AquaTru): $400–$700 unit + $120–$200 per year. Five-year envelope: about $1,000–$1,700.
- Whole-house RO: $1,500–$3,000 + $200–$400 per year. Five-year envelope: about $2,500–$5,000.
- Gravity (Big Berkey, Berkey Travel): $300–$500 unit + $50–$100 per year on long-life elements. Five-year envelope: about $550–$1,000.
Pitcher carbon is the cheapest by a wide margin. Gravity wins on cost per gallon if you stay in the system for years. Under-sink carbon and countertop RO converge in five-year cost — the choice between them depends on what you are trying to remove, not what you pay.
Water waste
Carbon and gravity produce zero waste. Every gallon you pour through becomes a gallon to drink.
RO is the exception. The EPA WaterSense overview notes typical POU RO systems waste around 5+ gallons per gallon delivered. WaterSense-labeled systems must hit 2.3 or lower, and a small but growing class of "tankless" RO claims close to 1:1. If municipal water cost is not your bottleneck, this matters less. If you live somewhere with strict drought-stage restrictions or very expensive water, the waste ratio is a real factor.
Mineral preservation and remineralization
Carbon and gravity preserve dissolved minerals in your water — calcium, magnesium, trace electrolytes. RO removes them along with contaminants. Many people find RO water tastes flat as a result. Most countertop and under-sink RO units, including AquaTru, offer optional remineralization cartridges to add minerals back.
The mineral question is also a health question for some readers. The WHO 2005 monograph on nutrients in drinking water reviewed long-term consumption of demineralized water and noted that for most people on adequate diets, food is the primary mineral source — drinking water is a minor contributor at best. The trade-off is largely a taste preference. If you do choose RO and dislike the taste, remineralize.
The decision framework — by priority
Match yourself to one row.
- "I just want my tap water to taste better." Pitcher carbon. Brita Elite for the cheapest entry; Epic Pure if you want a similar price point with PFAS testing.
- "I want comprehensive PFAS and heavy-metal removal, and I rent or cannot install." Countertop RO. AquaTru is the leading no-plumbing RO unit.
- "I own, want under-sink, and want filtration tuned to my local water." Custom carbon. Hydroviv builds the cartridge mix from your ZIP code.
- "My city uses chloramine, and I want under-sink convenience." Catalytic carbon. Aquasana AQ-5300 is built specifically for chloramine systems.
- "Off-grid living, emergency prep, no electricity required." Gravity. Big Berkey for stationary use; Berkey Travel for mobile.
- "I want the most aggressive pitcher claim I can buy." Premium pitcher carbon. Clearly Filtered has the broadest tested-contaminant list in the pitcher class. ZeroWater is the alternative with a TDS focus.
- "I want under-sink filtration I can mostly forget about." Crystal Quest Countertop is a multi-stage option that lives near the sink without a plumbing modification.
Whole-house catalytic carbon is the obvious next product class to add — it solves chloramine across every tap. None of the ten products in the current VettedClean catalog cover that use case, and we will note it explicitly when we add one.
What can mislead
"Removes 99.9% of contaminants" is the marketing line that should make you pause. The right follow-up: which contaminants, at what influent level, under what flow rate, and certified by whom. NSF/ANSI 42, 53, 58, 401, and P473 give you the contractual answers. Manufacturer-commissioned independent lab reports — what Berkey publishes, what Hydroviv publishes per ZIP code — are the second-best evidence.
Pitcher filters often appear to lose effectiveness faster than the cartridge schedule suggests. The carbon itself does not "wear out" so much as flow channels form, shortening contact time between water and media. The fix is replacing on time, not buying a different brand.
RO with remineralization is not strictly necessary for health. It is a taste preference. If you like the way unremineralized RO tastes, the WHO position implies you are not missing anything dietary by leaving it alone.
How to translate this into a buy
Start with what you are trying to remove, not the unit class. If lead and PFAS are your concerns, you are choosing between certified carbon and RO. If you live with municipal chloramine, you are choosing between catalytic carbon and RO. If you want bacteria-rated filtration without electricity, you are choosing gravity.
Then layer in install — renter versus owner, counter space, plumbing access. Then five-year cost. Water waste lands last, and only matters meaningfully for RO.
For a focused PFAS-only buying guide that compares specific products instead of mechanisms, see Best PFAS Water Filters in 2026. For the certification background on PFOA / PFOS testing, see What is NSF P473.
Frequently asked questions
(See structured FAQ at the bottom of this page for full answers.)
