Water softener vs water filter: which one your well test calls for

The short answer: a water softener removes hardness (calcium and magnesium) by swapping it for sodium or potassium. A water filter removes contaminants. Which one your well needs depends on the lab report, line by line.

That sounds obvious, and the trouble starts right after it. "Water filter" isn't one device. It's a family: sediment cartridges, activated carbon, reverse osmosis (RO), anion exchange, oxidizing media and more, and each one catches different things. A softener, by contrast, is one specific machine, a cation exchange unit, and the list of what it removes is short. The Minnesota Department of Health (MDH) says softening is designed to fully or partially remove "Calcium, copper, iron, magnesium, manganese, radium." Nitrate, arsenic, PFAS, solvents and bacteria aren't on that list.

So the real question is which line of your report you're trying to fix. Our hard water guide covers how to test hardness and convert the number. This page starts where that one stops: you have a report, and you need to know whether it points to a softener, a filter, both, or neither.

Water softener vs water filter: what each one actually is

Penn State Extension describes the softener's mechanism: resin "coated with positively charged sodium ions," and as hard water flows through, "the resin releases its sodium ions and readily trades them for the calcium and magnesium ions." When the resin is full, a brine regeneration flushes the hardness out. MDH adds that the exchange can be "for sodium or potassium." EPA's WaterSense program calls these units "cation exchange water softeners." That word, cation, is the key to everything below.

Filters work by other means. Wisconsin's health department (DHS) says carbon devices adsorb contaminants, "(stuck) onto a carbon filter," and reverse osmosis forces water through "a membrane (very thin material) with holes large enough for water molecules to pass through but small enough that they block most contaminants." Penn State adds a warning about the way softeners get sold: softening "has often been misconstrued as a purifying, cleansing or conditioning process," but "the reality is that water softening simply removes hardness minerals."

UV lamps and chlorinators don't filter anything; they disinfect, and they belong to the bacteria row further down.

Why a softener misses nitrate, arsenic and uranium: it's the electric charge

Most of what a softener can and can't do comes down to one fact. Its resin trades positive ions. Plenty of well contaminants dissolve as negative ions or carry no charge at all, and the resin has nothing to grab them with.

• Nitrate is a negative ion. Penn State explains that a nitrate unit "operates much like a household water softener," but "For nitrate removal, the resin exchanges chloride ions for nitrate and sulfate ions in the water." Different resin, opposite charge.
• Radium is positive, uranium is negative. Vermont says a softener "can be used to reduce the level of radioactive ions with a positive charge, like radium," but "This will not lower the levels of uranium in the water." Anion exchange works the other way: it "exchanges the negatively charged radioactive ions, like uranium for chloride," and "will not lower the level of radium."
• Arsenic depends on its form. EPA's arsenic treatment training explains that at the pH of natural water (typically 6 to 9), arsenate, or arsenic(V), "occurs as a charged particle," while arsenite, arsenic(III), "occurs as an uncharged particle." Since arsenite's net charge is neutral, "this form is not easily removed." EPA adds that groundwater supplies "typically contain high levels of arsenite," so pre-oxidation "will probably be necessary." MDH's treatment sheet says the same for Minnesota wells, where both forms occur. And even the negative form is the wrong charge for a softener, which fits Vermont's plain statement that a softener "will not remove arsenic."
• PFAS and solvents aren't a softener's job. MDH: softeners and iron filters "are not likely to remove PFAS," and "sampling data indicate that those systems do NOT remove PFAS." For volatile organic compounds (VOCs), the device tables from both MDH and Wisconsin DHS list carbon (MDH also lists RO and aeration). Neither table lists a softener.
• Bacteria can grow in a softener. A softener doesn't disinfect. MDH warns that if water "is not disinfected prior to softening, bacteria and fungi can potentially grow on the surface," and Penn State says they "form mats in the resin that reduce its effectiveness."
• Lead and copper are a plumbing problem. Penn State: "Contrary to some claims, water softeners are not recommended for lead removal," partly because softeners "usually are installed in the plumbing system ahead of the piping and fixtures where most of the lead originates." MDH says a softener may remove copper, but "copper can reenter the water" after it leaves the softener.

Wisconsin DHS's one-page device table makes the same point. Of the 12 substances it lists as commonly tested in Wisconsin wells, the softener column is ticked for one: manganese. Reverse osmosis is ticked for eight, carbon for seven.

The U.S. Geological Survey's national sample of private wells shows why this matters. Hard water turned up in about 60 percent of wells, the most common finding in the chart. But the same untreated samples turned up total coliform bacteria in 34 percent of the wells tested for it, and arsenic, nitrate and uranium above EPA's limits in a few percent each. Those are the results a softener leaves in the water. (EPA's limits are written for public water systems; they don't legally apply to a private well, but they are the usual yardstick.)

Read your report line by line: which result points to which device

Find each line of your report in the left column. Where a source sets a number, the number is in the row. The treatment-by-test-result guide has the full detail for each technology, including failure modes and the order units go in.

Three rows need numbers you may not have. Rows 2 and 3 need iron and pH, and row 3 needs to know whether the iron was dissolved when it left the well. Row 7 needs speciation, a separate test. New Hampshire's environmental agency (DES) recommends "that if water tests above 5 ppb for arsenic, a follow-up test for Arsenic-III be conducted before selecting a treatment system." Our guide on arsenic removal covers it, and the guide to reading a well water test report explains units such as mg/L, µg/L and pCi/L.

Two worked examples

Both reports are illustrations.

Example 1: "Hardness 205 mg/L, iron 1.5 mg/L, pH 7.2, coliform absent, nitrate 3 mg/L"

205 divided by 17.1 is about 12 grains per gallon. Nitrate is below 10 and bacteria are absent, so nothing on the health side calls for treatment. That puts the report in row 2. The numbers sit inside Penn State's window: pH above 6.7, hardness between 3 and 20 gpg, dissolved iron under 5 mg/L. A softener could handle both the scale and the iron, provided the water reaches it before it meets air or chlorine. If the sample came out of the tap orange, move to row 3 and plan a filter ahead of it.

The cost is sodium. At Penn State's rate of 7.5 mg per quart for each grain removed, which works out to about 7.9 mg per liter, 12 grains adds roughly 95 mg/L. Wisconsin DHS recommends a threshold of 20 mg/L for people on a low-sodium diet. If anyone in the house is on one, row 11 applies too, and the conversation belongs with their doctor.

Example 2: "Hardness 150 mg/L, nitrate 14 mg/L as N, arsenic 12 µg/L"

This is the report where the softener pitch goes wrong. The hardness is real (about 8.8 gpg, row 1), but the two health lines are rows 6 and 7. Nitrate is 1.4 times EPA's 10 mg/L limit and arsenic is 1.2 times EPA's 10 µg/L (both are limits for public water systems). A softener would change the scale and leave both untouched.

Both rows name reverse osmosis, and Vermont lists RO for arsenic and nitrate. Vermont also says to "Be sure any product you use is certified to remove the specific contaminants in your water," so the RO unit needs a claim for each one, not just an NSF/ANSI 58 mark. At 150 mg/L, the hardness sits inside the 120 to 170 mg/L range that Vermont gives as a general maximum ahead of an RO membrane, so check the manufacturer's hardness limit. If it is lower, a softener comes back in, as pretreatment for the RO rather than as the fix. At 12 µg/L (12 ppb), arsenic is over New Hampshire DES's 5 ppb trigger, so get the arsenic(III) test first. Then test the RO tap: Wisconsin DHS says to test treated water after installing a device "to make sure the device is working."

Water conditioner vs water softener

"Water conditioner" has no official definition in any of the agency or extension sources we read, and the word gets attached to different machines. EPA's May 2026 WaterSense guide uses "salt-free water conditioners" for devices that, "Instead of removing the hardness ions," condition them; that is a description, not a definition. So ask one question of any "conditioner": what does it do to the calcium and magnesium?

Nebraska Extension's December 2025 guide is the most thorough official review of the salt-free kind. It calls them "descalant devices as opposed to hardness removing devices, since they do not remove the hardness from the water." Template-assisted crystallization uses resin beads "to convert (not remove) dissolved hardness ions to microscopic scale-resistant crystals." It "typically require[s] relatively clean water" and may need pretreatment for high iron, manganese or sediment. On results, Nebraska says "few conclusive studies confirm the efficiency of these devices for residential use," and "until product testing for household anti-scaling products is available, consumers cannot tell for certain which products best reduce scale buildup in the home." EPA is warmer. Its 2026 guide says these units reduce "scale buildup without added chemicals or water waste" and tells buyers to "look for models that meet the requirements of the IAPMO/ANSI Z601 Standard for Scale Reduction Devices."

The test-first consequence is easy to miss. Our reading of Nebraska's "do not remove the hardness": because a descaler leaves the hardness in, a lab hardness test after installation should look about the same as before, so the report can't show it working the way it can for a softener. Electrochemical deionization units are the exception; Nebraska says they can treat water to "less than 5 grains per gallon," and a lab test would show that. And a descaler does nothing for any health row in the table above. Our hard water guide covers Nebraska's view of the electrochemical and magnetic types.

Water softener vs reverse osmosis

People compare these two because both change what's dissolved in the water. They do different jobs at different taps, and many wells that need one end up with both.

Vermont gives the main reason they show up together: "Hard water can cause scale on RO membranes and a pre-treatment system may be needed, such as a water softener," and in general hardness "should not be higher than 120 to 170 mg/L." The pairing has a second use. Sodium is on MDH's list for RO, so an RO tap at the kitchen would take out much of the sodium a softener adds to the drinking water. That's our reading of MDH's table, not an MDH recommendation, so test the RO tap and ask a doctor if sodium is a medical concern.

RO alone removes hardness only at the tap it serves; the water heater and showers stay hard.

Salt, sodium and the septic system

A softener's two side effects land in different places: sodium in your drinking water, chloride and extra water in your septic system. The hard water guide has the full sodium arithmetic. The short version from MDH: if someone in the house "has a history of high blood pressure, consult a doctor about drinking softened water," and you can "Have an un-softened tap for cooking and drinking" or regenerate with potassium chloride. Wisconsin DHS notes "There are no state or federal standards for the amount of sodium allowed in drinking water," and that "The largest sources of sodium in Wisconsin's drinking water are road salt and water softeners."

On septic systems, agencies agree about the water and disagree about the salt.

• EPA's homeowner septic guide (2012) says some systems, "including water softeners," can "send hundreds of gallons of water to septic tanks, causing agitation of solids and excess flow to drainfields," and suggests asking a licensed plumbing professional "about alternative routing."
• EPA's WaterSense guide (May 2026) says that because softeners discharge salt to the wastewater stream, "they can impact local water quality or septic system efficacy," that "some local jurisdictions prohibit the installation and use of these systems," and that septic owners should check with "the septic system manufacturer or a local service provider" before buying.
• Penn State estimates "about 50 gallons" per regeneration, which "may or may not cause hydraulic overload of the septic system."
• Virginia's health department took the opposite view on salt in a 1994 policy memo, still listed among its active onsite policies (index updated April 2025). Citing University of Wisconsin research done for the water treatment industry's trade association, it said brine "had no adverse effect on the bacteria in the tank" or on the soil, and that similar Virginia Tech work found the only concern was hydraulic overloading. Its policy: softeners shouldn't harm a septic system if a regeneration sends "60 gallons of brine or less" per cycle and "No more than 180 gallons" per week. Above that, the system design has to be enlarged.
• MDH points out where the salt ends up: with a private well, softener chloride "drains to the home's septic system and then ends up in lakes and streams."

Reverse osmosis isn't free of this either. Vermont says the reject water from an RO unit "is directed to the septic system or a drywell." What lowers the load from a softener is in the agency advice. Size it from a lab number (MDH says a setting that's too high will "cost more to operate and waste water"). Pick demand-initiated regeneration (EPA, Vermont). Soften only the water you need, often just the hot water (MDH, EPA). Before installing, ask your county health department whether a separate discharge is allowed or required where you live.

Before you buy either: test first, then test the treated tap

1. Get an independent lab report. Penn State: "Although many water softening companies offer free hardness testing, it's best to have a third-party evaluation." MDH lists, among deceptive sales practices, in-home "tests" that "may simply indicate the presence of naturally occurring minerals in the water."
2. Order the lines the table needs. Hardness, iron, manganese and pH for the softener decision, plus the health tests for your area. Our which-tests tool builds the list, and the lab finder lists state-certified labs. Vermont: "Before deciding on a water treatment system for your home, it is important to complete all the recommended drinking water testing."
3. Sample raw water. Our advice: take the sample from a tap that bypasses any existing softener or filter, or you're testing the equipment, not the well. The how-to-test guide covers sampling.
4. Match each flagged line to a row in the table above, and note which rows a softener doesn't touch.
5. Re-test after installing. Vermont says to re-test after any radium or uranium treatment "to make sure levels are below the drinking water standards," and Wisconsin DHS says the same for any device connected to your plumbing. Penn State notes that most softeners "are not rated for effectiveness," so your own before-and-after test is the check that counts.

If you already have a report, enter the numbers in our results checker to see which lines are flagged before you get quotes.

Questions people ask