Pool Water Testing: What to Test, How Often, and How to Read It

Testing is the least glamorous thing on a service visit and the one everything else depends on. Every dose you add, every decision to leave a pool alone, every conversation with an owner about why their water looks wrong — all of it rests on a set of numbers you took in about ninety seconds at the edge of the pool.

Which means a bad reading is not a small error. It is an error that propagates. A free chlorine number that is half a point off sends you home having done nothing to a pool that needed something, or having dosed a pool that did not. And the most common cause of a bad reading is not a faulty kit. It is a sample that was never representative of the pool in the first place.

What follows is what each reading is actually telling you, which testing method earns its place where, how to draw a sample that means something, how often testing genuinely needs to happen, and what to do on the days the number in front of you looks wrong.

The six readings, and what each one is actually telling you

Free chlorine

The portion of chlorine in the water still available to sanitise. Total chlorine is free chlorine plus the combined chlorine that has already reacted with contaminants and is largely spent. The difference between the two numbers is the part that tells you whether the pool is keeping up.

Most people who test their own pool look at one number and stop. On a route you are reading six, and they are not independent of each other — several of them exist mainly to explain why another one is misbehaving.

Free chlorine is the working number: what is left to do the sanitising. Combined chlorine is what has already been used up reacting with sweat, sunscreen, leaves and everything else that goes into a pool, and a rising combined reading is the classic signal of a pool that is being asked to handle more than it can. The smell people call "too much chlorine" is almost always the opposite — combined chlorine that has built up because there was never enough free chlorine to burn it off.

pH is the one that quietly decides whether anything else works. Chlorine gets steadily less effective as pH climbs, so a pool sitting high can have a perfectly respectable chlorine reading and still be losing ground. It also drives comfort and corrosion at both ends: low pH eats plaster and equipment, high pH leaves scale and cloudy water.

Total alkalinity is pH's shock absorber. When it is too low, pH swings wildly on very little provocation and nothing you do to it holds; when it is too high, pH drifts upward and resists correction. If you find yourself adjusting pH at a property every single week, alkalinity is usually the reason, and fixing the pH again is treating the symptom.

Calcium hardness matters at both extremes and almost nowhere in between. Too low and the water will pull calcium out of the plaster to get what it needs; too high and it deposits it back as scale on surfaces and inside heaters. Cyanuric acid — stabiliser — is the one that catches people out, because it does something genuinely useful and then keeps doing it after you would rather it stopped. It shields chlorine from sunlight, and it also holds chlorine back from working, so a pool with stabiliser that has climbed season after season can show a healthy chlorine reading while very little of that chlorine is active.

Read together these six usually name the problem before you have looked at the pool properly. A cloudy pool with high pH and high calcium is a different job from a cloudy pool with a clean chemistry panel, and working out which kind of cloudy you are looking at starts here rather than at the filter.

Strips, drops or digital

There is a long-running argument about test strips that is mostly people talking past each other. Strips are not inaccurate so much as coarse, and coarse is fine for the job most of them are doing. The mistake is not using strips. The mistake is using a strip to make a decision that needed a better number, and never noticing that you did.

Taking a sample that means anything

This is the part that separates a number from a guess, and it is almost never the part that gets taught. A sample scooped off the surface next to the skimmer is not a sample of the pool. It is a sample of the one spot where everything the pool is trying to get rid of is currently collecting, taken from the layer most affected by sunlight and the last thing anybody poured in.

1. 1

   Test before you do anything else to the pool

   Not after brushing, not after emptying baskets, and certainly not after adding anything. The reading you want is of the pool as you found it, because that is the pool that has been running unattended since your last visit. Anything you do first changes the thing you are trying to measure.

2. 2

   Get away from the skimmer and the returns

   The skimmer throat concentrates debris and surface film. A return jet is delivering water that has just come through the filter and past anything dosing into the line, so it reads better than the pool does. Stand somewhere between the two, ideally at the deep end, well away from both.

3. 3

   Draw from depth, not from the surface

   Push the vial in upside down, take it down about elbow-deep, then turn it over and let it fill there. The top few inches of a pool sit in direct sun, hold whatever blew in most recently and are the least representative water in the whole body. Turning the vial over at depth is a two-second habit that removes a whole class of wrong answers.

4. 4

   Use a clean vial

   Residue from the last test, a film of sunscreen off your own hands, or a vial that dried with reagent still in it will all shift a reading. Rinse it in the pool you are about to test, not in the last one, and keep fingers off the inside and off the optical faces of a photometer cell.

5. 5

   Test it immediately, at the pool

   A sample starts changing the moment it leaves the water. Chlorine falls off in sunlight and warmth, so a vial that rides on the dashboard to the next stop is testing something other than the pool it came from. If a sample genuinely has to travel, keep it capped, full to the top and out of the light, and treat the result as indicative rather than final.

6. 6

   Read the colour in shade, against something neutral

   Direct sun, a coloured deck, a shirt and sunglasses all shift how a colour reads, and every one of those pushes the judgement in a different direction on a different day. Hold the comparator against open sky or a white surface, in shade, and take sunglasses off. If two technicians on the same route read the same pool differently, this is usually why.

7. 7

   Write it down before you dose

   Record the reading first, then decide, then record what you added. Doing it in that order means the history shows what you were responding to rather than what you did, which is the difference between a log you can diagnose from and a list of chemicals.

None of that takes meaningfully longer than doing it carelessly. It is perhaps thirty seconds of difference across a stop, and it is the thirty seconds that decides whether the rest of the visit was based on anything real.

How often, and what it depends on

The honest answer to how often a pool should be tested is that every reading is only good until something changes, and the useful question is what counts as something. For a serviced residential pool on a weekly route, the baseline is straightforward: the full panel every visit, because the visit is the only time anybody is looking. What varies is everything sitting on top of that baseline.

Two readings do not need that frequency, and testing them every visit is mostly wasted reagent. Calcium hardness and cyanuric acid both move slowly and only really change when water is added or replaced, so monthly is usually enough — with the caveat that stabiliser is the one people forget for whole seasons, and it is the one most likely to be quietly causing the problem you are chasing.

When you do not believe the reading

Every technician has had the moment where the number in front of them does not match the pool in front of them. Water that looks fine reads as though it has no sanitiser in it; a pool that has visibly gone over reads perfectly balanced. The instinct is to trust your eyes and dose to what the pool looks like. That instinct is usually wrong, and acting on it is how you end up over-dosing a pool that was fine.

There is also a category of pool where the readings are genuinely fine and the problem is real anyway. Black algae is the clearest example: a pool can test perfectly and still have a live colony on the wall, because the reading describes the water and the colony is not in the water. That is the opposite situation from a pool that has turned green, where the readings and the problem agree with each other completely. Knowing which of those you are standing in front of is the whole diagnosis.

What disciplined testing is worth on a route

On one pool, careless testing costs a bag of chemicals. On forty pools it compounds into something else entirely: chemical spend that nobody can account for, call-backs that arrive without explanation, and a service history that cannot answer the only question that ever gets asked, which is what changed.

The trend is the part that is hard to see from inside a single visit and obvious from a season of records. A pool whose alkalinity has been drifting down for two months is telling you something a Tuesday-afternoon reading cannot. So is a pool whose chlorine demand has crept up week after week while nothing else moved — which is frequently a filter that has stopped doing its job rather than a chemistry problem at all.

The record is half the value

A reading that is taken and not written down has done half a job. It has told you what to do today and nothing about what is happening to the pool, and the second half is where the money is — in catching the slow drift, in defending a chemical charge, and in handing a route to a technician who was not there last month.

That record is also what turns a chemical charge from an assertion into a demonstration. An owner questioning what they were billed for is a reasonable question, and the answer is much easier when it is a run of dated readings showing demand climbing rather than a general statement that the pool needed it. It is one of the practical reasons what a service agreement says about chemicals and testing is worth settling in writing before the season rather than during it.

None of this is difficult. It is a sample taken properly, a kit that has not cooked on a truck, a reading written down before a decision is made, and the discipline to test twice when something looks wrong instead of dosing once and hoping. Done consistently it removes most of the chemistry problems a route ever has — not by fixing them, but by finding them while they are still small enough to be uninteresting.