Koi Pond Water Testing: What the Numbers Actually Mean

Koi pond water testing is the part of the hobby nobody puts on Instagram, and it's the part that decides whether your fish are still here next spring. We run a koi shop in Miami, we keep a lot of fish in our own tanks, and we get called out to ponds all over South Florida — and the pattern almost never changes. When a pond goes wrong, the water went wrong first, usually days earlier, and usually in a way a test kit would have caught for the price of about thirty seconds.

So here's how we actually read the numbers, what each one is telling you, and the handful of readings that make us stop what we're doing.

The numbers that matter most in a koi pond

Ignore the fourteen-parameter test strips for a second. For a koi pond, you are watching ammonia, nitrite, pH, and KH (carbonate hardness, sometimes labelled alkalinity). Ammonia and nitrite should both sit at or near zero, always. pH should be steady somewhere in the 7.5 to 8.3 range — the word "steady" is doing more work in that sentence than the numbers are. KH we want at 120 to 180 ppm in an established koi pond, or 100 to 150 ppm while a new pond is still cycling.

Anything below about 80 ppm KH is a red flag — that's the reading that makes us pick up the phone.

Why we check KH before we check pH

This is the mechanism most pond keepers have never had explained to them, and once you see it you can't unsee it.

The bacteria in your filter that convert ammonia into nitrite and then nitrate don't do it for free. Nitrification consumes roughly 7.1 mg of alkalinity for every 1 mg of ammonia-nitrogen it processes. That's the whole trick. Heavy feeding plus warm water plus a decent fish load means your filter is working hard, and every bit of that work is quietly eating your KH.

So KH trends down for days. Nothing looks wrong. pH drifts a little, and you tell yourself it's fine. Then the buffer runs out, and pH doesn't drift any more — it falls off a cliff, overnight, and you walk out in the morning to a pond that looks nothing like the one you left. We've seen this exact sequence more times than we'd like, and the owner almost always says the same thing: "but it was fine yesterday." It genuinely was. That's the problem with a buffer — it holds, and holds, and then it doesn't.

Test KH once a day if you're feeding hard in warm water. It's the early warning for a pH crash that hasn't happened yet.

Nitrite: the number that means your filter is behind

Ammonia gets all the attention and nitrite gets forgotten, which is exactly why it catches people out. It's the middle step. One group of bacteria turns ammonia into nitrite, and a second group turns nitrite into relatively harmless nitrate. If that second population isn't established yet, or it's been knocked back, ammonia clears and nitrite piles up behind it.

So nitrite turns up in the situations you'd least expect. A new pond that looks like it finished cycling. A filter that got cleaned a bit too thoroughly. A pond that's been medicated. One where feeding jumped faster than the bacteria could follow. Your ammonia can read a clean zero while nitrite is climbing.

And it's a gill problem. Nitrite crosses at the gills and interferes with the blood's ability to carry oxygen, which means a fish in high nitrite looks exactly like a fish in an oxygen crash — gasping at the surface, in water that tests fine for oxygen. If you're chasing an oxygen problem you can't find, test nitrite.

The useful part is that you can buy time. Chloride competes with nitrite for uptake at the gills, so salt at 0.10 to 0.15% blocks much of the damage while the filter catches up. That's support, not a cure — the actual fix is to stop feeding, get the biofilter working again, and change water. A dechlorinator that binds ammonia as well as chlorine (Seachem Prime or Safe, ChlorAm-X, Amquel) is worth keeping on the shelf for exactly this situation.

pH, ammonia and nitrite are the three we run on every pond we visit. KH is the one we add when we want to know what's coming rather than what already happened.

Your midday pH reading is lying to you

Here's the one that catches nearly everybody. Plants and algae strip CO2 out of the water all day through photosynthesis, and then respiration puts it back overnight. CO2 drives pH. So the same pond can read 8.2 at four in the afternoon and sit in the 6s by sunrise.

If you test at lunchtime because that's when you're home, you are testing the pond at its most flattering moment. A clean midday reading means almost nothing on its own.

What we tell people: log pH twice — once at sunrise, once at 4pm — and log KH once a day. Two weeks of that tells you more about your pond than a year of random midday spot checks. The gap between those two pH numbers is your swing, and a big swing means your buffer is thin even if neither individual number looks alarming.

Oxygen follows the same daily curve, and it bottoms out just before dawn for the same reason. Which brings us to the emergency.

What an oxygen crash actually looks like

People send us photos of one fish sitting near the waterfall and ask if it's an oxygen problem. Usually it isn't — one unhappy fish is usually one unhappy fish.

An oxygen crash hits everything at once. That's the signature. All the fish affected simultaneously, gasping at the surface, and in bad cases holding the mouth open. Gills that should be red look brown. Deaths happen overnight or in the early morning, not in the afternoon. There's usually a heavy algae or organic load in the background, or an algae die-off, or a bacterial bloom, or it's just been very hot.

A mouth stuck open is not an early sign. That is late-stage, and it means move now.

The other thing worth saying plainly, because it comes up on service calls constantly: a skimmer moving the surface is not aeration. Surface movement looks like aeration. It isn't. A heavily stocked pond needs dedicated aeration, and if you're in a part of the world where the power goes out in summer storms — hello, South Florida — it needs a backup too. A pond can lose its fish in a single overnight outage.

The South Florida problems that don't show up in the books

Most pond literature is written for climates that get a real winter, and a lot of it doesn't transfer. Three things we run into here constantly:

Hard water is not pH insurance. Miami water is hard and people assume that means their buffer is bulletproof. GH and KH are two different values measuring two different things, and you can absolutely have hard water and a KH crash in the same pond. Test KH specifically. Don't infer it.

Rain is our version of a cold snap. The seasonal hazard here isn't freezing — it's runoff. A serious downpour carries mud, lawn fertiliser and pesticide straight into a pond that sits at the low point of the yard, and it dilutes your buffer at the same time. If you want one habit out of this whole article: test KH the morning after heavy rain. Leaves and seed pods also pack skimmer baskets faster than people expect, which is why bi-weekly attention is realistically the floor for a South Florida pond, not the luxury option.

Hard water quietly kills your UV. Calcium builds up on the quartz sleeve, the lamp still glows, the unit still hums, and the UV is doing essentially nothing. It fails while appearing to run, which is the worst way for equipment to fail. The same hard water seals the pores of ceramic filter media — plastic moving-bed media or matting holds up much better here.

What to do when a number goes wrong

Stop feeding. That's the first move for almost any water-quality instability, and 24 to 48 hours without food costs a koi nothing. Koi are stomachless cyprinids — they're built for small frequent meals, not for one big one — so cutting food immediately cuts the ammonia load going into a filter that's already struggling. Normal feeding is whatever they clear in about five minutes, but when the water is unstable the right amount is none.

Then fix the actual number rather than chasing symptoms, and resist the urge to correct pH directly. pH is usually the readout, not the disease; KH is generally the thing to address.

And if the fish that triggered all this is a new arrival, this is the moment to mention that new koi belong in quarantine for at least 30 days before they ever see your main pond — we wrote a full walkthrough in How to Quarantine New Koi. Most of the "mystery" outbreaks we get called about trace back to a fish that went straight into the pond. As the water cools later in the year, portion size matters more than people expect too, which we covered in feeding koi in late summer and into fall.

None of this is complicated. It's a test kit, five minutes a day, and writing the numbers down somewhere you'll actually look at them. The pond keepers who never have a crisis aren't luckier than everyone else — they just noticed the KH sliding two weeks ago.

If you're in the area, come by the tanks and bring a water sample. We'll look at it with you.

Ready to add a fish to a pond you trust? Browse our Available Koi — every one photographed and filmed individually, so the koi in the video is the koi that ships.