Well Water Testing: What to Test For, What It Costs, How Often

Nobody is legally required to check your well. Public utilities answer to the Safe Drinking Water Act, so they sample on a schedule, publish what they find, and have to notify you when they fail. Private wells sit outside all of it. Whatever the aquifer is doing this year, you are the only person who will ever find out, and only if you pay for the test. The good news is that the baseline is cheap and the arithmetic is simple. The bad news is that most owners test once at closing and never again.

Why nobody tests your well but you

The Safe Drinking Water Act regulates public water systems. The usual threshold is a system serving at least 25 people or having at least 15 service connections, which a single-family well never reaches. Maximum contaminant levels, the total coliform rule, the required sampling frequency, the public notice when a system fails, none of that reaches you. Your state or county may require a test at property transfer, and most require one on a new well before the health department signs off. After that, silence.

That silence is the problem. Groundwater is not a fixed thing. A septic drainfield fails two lots uphill, a farm switches crops, a drought drops the water table and the pump starts drawing from a different zone. Nothing about your tap will tell you. Nitrate has no taste. Arsenic has no color. E. coli does not smell. The contaminants that make people sick are almost never the ones anybody notices, and the ones you do notice, the iron staining and the sulfur odor and the cloudy glass, mostly cost you money and laundry.

One more thing, because it catches people. The test your lender ordered at closing is not a baseline. It was one sample, on one day, in one season, run for whatever the loan required, and it belonged to the transaction rather than to you. Dig out the actual report if you can still get a copy. If nobody has one, assume you are starting from zero.

The annual baseline every well owner needs

Six parameters, once a year. This short list either tells you the well is fine or tells you exactly where to look next.

• Total coliform and E. coli. Coliform bacteria are the indicator, not the danger. Their presence means surface water or something else is reaching the aquifer or getting past the well casing. E. coli means fecal contamination specifically, and that is a stop-drinking-now result.
• Nitrate. The EPA MCL is 10 mg/L as nitrogen. Fertilizer, manure and failing septic systems are the usual sources, and shallow wells see it first.
• pH. Not a health number on its own. Below about 6.5 the water turns corrosive and starts pulling copper and lead out of your own plumbing, which makes it a health number by a different route.
• Total dissolved solids. A general gauge of mineral load, and a useful year-over-year trend line. A jump in TDS with no other explanation is worth chasing down.
• Hardness. Reported in mg/L as calcium carbonate. Divide by 17.1 for grains per gallon, which is the number every softener on the market is sized on.
• Iron and manganese. Secondary limits of 0.3 mg/L and 0.05 mg/L. Above those you get orange or black staining in fixtures and laundry.

That iron line is where sources start arguing with each other. Some softener manufacturers rate their valves for two or three mg/L of clear-water iron as long as you run a resin cleaner on a schedule. Plenty of well drillers will tell you anything over 0.3 mg/L belongs on a dedicated filter. Both camps are describing something real. Dissolved ferrous iron passes through the resin and gets stripped like any other cation, while iron that has already met oxygen shows up as rust particles that coat the beads and stay there. A standard lab report gives you one iron number and does not separate the two forms, so on a well above 0.3 mg/L the safer read is to price a filter ahead of the softener and let the softener handle hardness only.

Hardness and iron together are also where testing stops being a health chore and turns into a purchasing decision. Grains per gallon is the number every dealer will quote you against, so that single line on the report is most of what you need to work out the grain capacity the house actually calls for.

lab says     = 205 mg/L as CaCO3
grains       = 205 / 17.1 = 12.0 gpg
USGS class   = very hard (over 10.5)
iron, same report = 0.9 mg/L
0.9 > 0.3 -> iron filter, then softener
size the softener on 12.0 gpg

Situational tests and who actually needs them

Running every available analyte every year is a waste of money, and most well owners never need most of this table. These are driven by geology and by what happens on the land nearby. Test once, then repeat on a long cycle only if the first result landed anywhere near a limit.

If you add one thing to the baseline, make it arsenic. It occurs naturally in bedrock, it turns up across large parts of New England and the Upper Midwest, and it has no taste or color at any concentration you would ever drink. Treatment for it also looks nothing like treatment for anything else on your report. Most state geological surveys and health departments publish a county-level map of where arsenic shows up in wells. Look at that map before deciding you live somewhere it does not.

PFAS is where the genuine disagreement lives. The EPA finalized federal MCLs for several PFAS compounds in 2024, with PFOA and PFOS set in the low single digits of parts per trillion. Several states had already published their own guidance values, and those numbers do not match each other, some stricter and some looser. Neither side is being careless. The health-based modeling depends on assumptions about how long exposure lasts and which effect you are protecting against, and different agencies made different defensible calls at different times, on different evidence. The federal rule has also been revisited and litigated since it landed, so confirm the current number with your state program rather than trusting any article, this one included. If your result falls between your state value and the federal one, you are in a real gray area, not a clerical error. Worth remembering too: those MCLs bind public systems. For your well they are a yardstick, not a rule.

Certified lab or home strip kit

Both have a place. The mistake is using the cheap one for the decision that needs the expensive one.

A state-certified lab gives you a defensible number with a detection limit printed beside it. Start with your county health department. Many run a lab themselves, most keep a list of certified ones, and some subsidize bacteria testing outright. Certification matters more than the brand on the letterhead, because a certified lab has to prove its methods against known samples on a schedule and gets audited on the results. Ask about turnaround before you collect anything. Bacteria results usually come back in a few business days, metals and organics in one to two weeks, and a bottle you fill on a Friday afternoon is often a bottle you have wasted.

What each tier of testing costs

Prices move with the state you are in, with whether the lab is public or private, and with how many analytes can share a single sample prep. A county health lab running coliform for $25 and a commercial lab charging $60 for the same test are both being honest. They have different cost structures behind the same result.

Over ten years, a yearly baseline plus one broad panel comes to somewhere between $850 and $2,000 depending which end of each range you land on. Set that against the equipment those numbers are supposed to be choosing. I have watched people get quoted for whole-house treatment on the strength of a free in-home test run by the company selling the equipment, sometimes with a little theater around the color of the sample. That test is a sales tool. It is usually hardness and TDS, and it tells you nothing about nitrate, arsenic or bacteria. Pay for your own numbers first. Comparing a filter, a softener and reverse osmosis only means anything with a lab report sitting on the table.

How to take the sample without ruining it

A bad sample costs you the fee and hands you a false positive that sends you chlorinating a well that was never sick. Plenty of surprise coliform hits come from the faucet and the technique, not from the aquifer.

• Use the lab’s own bottle. The bacteria bottle contains sodium thiosulfate to neutralize chlorine. Do not rinse it out, and do not touch the inside of the cap.
• Pick a tap ahead of any treatment equipment, ideally an outside spigot or the one closest to the pressure tank, unless you are deliberately testing treated water.
• Remove the aerator and take off any hose. Screens harbor bacteria that have nothing to do with your aquifer.
• Disinfect the spout with a flame or a bleach wipe, then run cold water three to five minutes so you are sampling the well and not the pipe.
• Reduce to a pencil-thin stream, fill to the line, cap it immediately.
• Chill it and get it to the lab fast. Holding time for a potable bacteria sample runs about 30 hours from collection. Sample on a Monday, not a Friday.

I am not a licensed plumber or a certified sampler. If a lender or a health department needs the result, ask whether they require a third-party collector, because some do, and a DIY sample rejected on chain-of-custody grounds is a wasted week.

Reading the report against EPA limits

Three columns matter: your result, the units, and the detection limit. Units trip people constantly. Nitrate is normally reported as mg/L as nitrogen, arsenic in micrograms per liter, and micrograms per liter is the same thing as parts per billion. Ten micrograms per liter of arsenic is 0.010 mg/L. Read those two as the same scale and you are off by a factor of a thousand. That is the most common misreading of a water report there is, and it goes wrong in both directions, panic as often as false comfort.

Compare your numbers against the EPA’s primary standards, which are the enforceable health-based MCLs, and its secondary standards, which cover taste, odor and staining and are not enforceable anywhere. Iron, manganese, TDS and pH sit in that second group. A result over a secondary standard is a comfort and equipment problem. A result over a primary standard is a health problem, and the two deserve very different amounts of your money.

“Non-detect” means below what the method can see, not zero. If your PFAS detection limit sits right at the federal MCL, a non-detect on that line is far weaker reassurance than the same word next to nitrate, where the limit is thousands of times above what the instrument can find. When you start shopping, match the certification to the analyte: NSF/ANSI 53 covers health contaminant reduction, 58 covers reverse osmosis systems, 44 covers cation-exchange softeners, 55 covers UV, 42 is aesthetic only, and 401 covers emerging contaminants. A device certified to 42 alone does nothing for arsenic, whatever the box implies.

If coliform comes back positive, resample before you do anything drastic. If it is positive twice, or E. coli appears even once, stop drinking it, call your county health department, and read up on the difference between shock chlorination and permanent UV disinfection before someone sells you one or the other. Shock chlorination handles a one-time event. It does nothing for a well that keeps getting reinfected, and a certified lab plus your health department will find the cause faster than any equipment quote will.

Events that should trigger an off-schedule test

The calendar is the floor, not the whole schedule. Test again when any of these happen.

• Flooding. If water covered the wellhead, assume contamination until a bacteria test says otherwise. Do not drink it in the meantime.
• Any work on the well. New pump, new casing, pulled pipe. Opening a well introduces whatever was living on the equipment.
• A pregnancy or a baby in the house. Test nitrate before the baby arrives and again a few months in. Infants under six months are the reason the 10 mg/L limit sits where it does, and formula mixed with high-nitrate water is the specific risk. Ask your pediatrician about it, not a water dealer.
• New land use nearby. A feedlot, a new septic field, a buried fuel tank, quarry blasting, or a neighbor drilling deep enough to change local drawdown.
• A change in taste, color or smell. Usually aesthetic. Sometimes not. A sulfur odor that showed up last month has its own short list of causes, and none of them get diagnosed by guessing.
• Recurrent stomach illness in the household with no other explanation.
• Drought, or a very wet spring. Both change what the aquifer is drawing in.

If you are on a septic system as well, the two are linked far more tightly than most owners realize, and keeping the tank on a real pumping schedule is part of protecting your own drinking water. A drainfield that surfaces is a bacteria source sitting on your own property, uphill of your own well.

Keep the reports. Ten years of results in one folder is worth more than any single test, because a trend tells you things a snapshot cannot. Nitrate creeping from 3 to 6 to 8 mg/L across four years is a story about something changing upgradient of you. A one-time 8 is just a number.