Is 0.25 ppm ammonia dangerous?
It depends on the pH, and the honest answer spans "no" to "yes" for the same number on the same kit. A reading is total ammonia: a mixture of NH₃, the un-ionized form that crosses the gill and does the damage, and NH₄⁺, ammonium, which mostly does not. Alkaline water pushes the mixture toward NH₃; warm water pushes it a little further. So the 0.25 ppm that sends people to a search box is:
- At pH 7.0 in a community tank at 25 °C: 0.56 % of it is NH₃ — 0.0012 mg/L as NH₃-N, under the long-term exposure limit — still fix it. The kit could read 4.3 ppm before the long-term exposure limit.
- At pH 8.2 — a Rift lake tank, or salt water: 8.2 % is NH₃, 0.017 mg/L, still just under the limit but 85 % of the budget spent. Another quarter of a ppm and it is over.
- At pH 8.5, the top of the Rift lake band: 15.2 % is NH₃, 0.0313 mg/L — over the long-term exposure limit. The tank needs a 36 % water change to get back under it, today.
And a full 1 ppm at pH 8.2 is 0.068 mg/L — over the acute threshold — change water now. The thresholds are the aquaculture ones: 0.02 mg/L NH₃-N as the long-term exposure guideline, 0.05 where acute damage starts. They are quoted as nitrogen, which is the first thing this page has to be careful about, below. Why the chart every other site prints says "0" and stops is the freshwater parameters guide's argument; this page is the tool that argument asks for.
Free ammonia chart: how much of a reading is toxic, by pH and temperature
The table is the split itself — the share of any total reading that is NH₃ — at five freshwater temperatures and in seawater, plus the column no hobby chart carries: the total ammonia a kit could read, at that pH, before free NH₃ reaches the long-term limit. Multiply your reading by the share; or read the budget column and compare it with your reading directly.
| pH | 18 °C | 22 °C | 25 °C | 28 °C | 30 °C | Sea, 25 °C | Budget, 25 °C |
|---|---|---|---|---|---|---|---|
| 6.0 | 0.03 % | 0.05 % | 0.06 % | 0.07 % | 0.08 % | 0.04 % | 43 ppm |
| 6.2 | 0.05 % | 0.07 % | 0.09 % | 0.11 % | 0.13 % | 0.07 % | 27 ppm |
| 6.4 | 0.09 % | 0.11 % | 0.14 % | 0.18 % | 0.20 % | 0.11 % | 17 ppm |
| 6.6 | 0.14 % | 0.18 % | 0.23 % | 0.28 % | 0.32 % | 0.18 % | 11 ppm |
| 6.8 | 0.22 % | 0.29 % | 0.36 % | 0.44 % | 0.50 % | 0.28 % | 6.8 ppm |
| 7.0 | 0.34 % | 0.46 % | 0.56 % | 0.69 % | 0.80 % | 0.44 % | 4.3 ppm |
| 7.2 | 0.54 % | 0.72 % | 0.89 % | 1.1 % | 1.3 % | 0.70 % | 2.7 ppm |
| 7.4 | 0.85 % | 1.1 % | 1.4 % | 1.7 % | 2.0 % | 1.1 % | 1.7 ppm |
| 7.6 | 1.3 % | 1.8 % | 2.2 % | 2.7 % | 3.1 % | 1.7 % | 1.1 ppm |
| 7.8 | 2.1 % | 2.8 % | 3.5 % | 4.2 % | 4.8 % | 2.7 % | 0.7 ppm |
| 8.0 | 3.3 % | 4.4 % | 5.4 % | 6.5 % | 7.4 % | 4.2 % | 0.45 ppm |
| 8.2 | 5.1 % | 6.8 % | 8.2 % | 10.0 % | 11.3 % | 6.6 % | 0.29 ppm |
| 8.4 | 7.9 % | 10.3 % | 12.5 % | 14.9 % | 16.8 % | 10.0 % | 0.2 ppm |
| 8.6 | 12.0 % | 15.4 % | 18.4 % | 21.8 % | 24.2 % | 15.0 % | 0.13 ppm |
| 8.8 | 17.7 % | 22.4 % | 26.4 % | 30.6 % | 33.6 % | 21.9 % | 0.09 ppm |
| 9.0 | 25.5 % | 31.4 % | 36.2 % | 41.2 % | 44.5 % | 30.7 % | 0.07 ppm |
Share of total ammonia present as NH₃ from Emerson et al. (1975): pKa = 0.09018 + 2729.92/T(K); fraction = 1/(1 + 10^(pKa − pH)). Sea column at 35 ppt adds 0.147 × ionic strength (0.72 mol/kg) to the pKa (Bell et al. 2007). Budget = total ammonia, as NH₃ on the kit, at which free NH₃ reaches 0.02 mg/L NH₃-N. Acute harm (0.05) is 2.5× the budget.
Does my kit read NH₃, NH₃-N or NH₄⁺? The 22 % nobody mentions
One quantity, three bases. A kit can report ammonia as the whole NH₃ molecule, as the nitrogen atom alone (NH₃-N, the convention in every toxicity study), or as the ammonium ion NH₄⁺. The same water reads 0.5 ppm as NH₃, 0.41 as NH₃-N, or 0.53 as NH₄⁺ — 22 % and 29 % apart from the nitrogen figure — and the thresholds above are in nitrogen. Every free-ammonia calculator we found silently assumes one basis; the fishless cycling calculator makes the same point about dosing. Here it is about reading.
| Kit or report | Basis, and what to do |
|---|---|
| API, Tetra, Aquarium Co-Op strips, Red Sea, Salifert | Total ammonia, as NH₃ — the card says "NH₃/NH₄⁺" and rarely states a basis. Enter it as NH₃. |
| Hanna checkers (HI700, HI715), Hach, most lab reports | As nitrogen — printed "NH₃-N" or "ammonia nitrogen". Enter it as NH₃-N; it is 22 % more ammonia than the same number as NH₃. |
| sera NH₄/NH₃, JBL NH₄ | As ammonium, "mg/L NH₄" — with a pH table on the leaflet. Enter it as NH₄⁺; this page is that table with temperature and salinity added. |
| Seachem Ammonia Alert, MultiTest "free" reading | Already free NH₃, not total. Compare it straight against the thresholds below (as NH₃, ×1/1.216 for NH₃-N); this calculator wants the total reading. |
Typical of current versions — the card or leaflet wins if it says otherwise. A JBL or sera "NH₄" reading converts to NH₃ units at ×0.944; the tool does this for you.
Enter it as NH₃-N. If the kit was really reading as NH₃ you have overstated the ammonia by 22 %, which errs toward a water change; guess the other way and you have understated it by 18 %. Given the choice of being wrong, be wrong in the direction that changes water.
How much water do I change to bring ammonia down?
Dilution, and only dilution: new water carries no ammonia, so replacing a fraction f of the tank leaves (1 − f) of the reading. The tool takes your reading and the budget at your pH and prints the fraction between them — 36 % for 0.25 ppm at pH 8.5 — and, with the tank volume filled in, the liters or gallons. Two things the arithmetic hides:
- A water change can make the reading more toxic. If the new water is harder or more alkaline than the tank — RO-blend soft water topped up with tap, an old acidic tank refilled with fresh — the pH rises, and every 0.3 of a pH unit turns roughly 1.8× more of what remains into NH₃. The total falls; the toxic share climbs faster. Match the pH, or change in smaller steps and retest. This is the mechanism behind the old advice not to do a huge change into a crashed-pH tank with ammonia in it.
- The budget is a ceiling, not a target. The long-term figure is where chronic effects are not expected; the tool sizes the change to get under it because that is the number a kit can act on. Getting to zero is the filter's job, and the nitrogen cycle guide covers why it can't be rushed.
A dechlorinator that "detoxifies ammonia" buys 24–48 hours by converting NH₃ to a bound form the kit may still read — the water conditioner page explains what it does and doesn't do to the number. It is the right thing to add after the change, not instead of it.
Salt water: the same reading, about a fifth less toxic
Dissolved salt raises the ammonium pKa, so at the same pH less of the total is NH₃. At 35 ppt and 26 °C the shift is 0.11 pH units of pKa, which makes any reading at pH 8.2 0.80× what the freshwater equation would say: the long-term budget is 0.35 ppm (as NH₃) instead of 0.28. Not a big correction — marine pH is what makes an ammonia reading in a reef or fish-only tank urgent, and salt only takes the edge off — but it is the direction most tools get wrong.
The freshwater equation (Emerson 1975) is settled. The seawater one is not: the expression most calculators copy, from Khoo et al. (1977), was shown in 2007 to be misprinted and to overstate NH₃ by up to 500 % in cold water, and the two replacements its correctors recommend disagree with each other by up to 23 %. This page uses the ionic-strength term from that 2007 re-fit on top of Emerson, which reproduces the Bower and Bidwell (1978) tables within a tenth of a percentage point across their salinity range. Read the salt-water result as a good estimate, not a measurement.
Planted tanks with CO₂: your reading changes twice a day
Injected CO₂ is carbonic acid. At 4 dKH the water sits at about pH 6.7 with 30 ppm of CO₂ in it at the end of the photoperiod, and drifts back up to about pH 7.7 by morning once the gas has been off all night. Same water, same total ammonia — and the 0.25 ppm reading is 9.8× more toxic just before lights-on (0.0053 mg/L NH₃-N) than at lights-off (0.0005). Two consequences:
- Enter the pH from the time of day you are worried about, not the one on last week's log. The planted option above takes your KH and prints both ends.
- A CO₂ system that fails off — an empty cylinder, a stuck solenoid — raises the pH by a point over a few hours, which is the same as multiplying any ammonia reading by ten. A cycling planted tank is the case to watch. The CO₂ calculator owns the pH-for-KH arithmetic this uses.
What to do at each level
| Verdict | What it means, and what to do |
|---|---|
| trace — under the long-term exposure limit — still fix it | Under the long-term limit. Stop feeding for a day, find the source (a death, overfeeding, a new tank, a filter clean), retest tomorrow. No emergency change — but if pH could rise (a big change with harder tap water, a CO₂ shutoff), the reading is worth more than it looks. |
| chronic — over the long-term exposure limit | Gill and growth damage over days to weeks. Change the fraction the tool prints today, dose a conditioner that binds ammonia for the next 24–48 h, feed nothing, retest in the morning and repeat until it reads under the budget. |
| acute — over the acute threshold — change water now | Damage is happening now. Change the printed fraction immediately with temperature-matched, dechlorinated water at the same or lower pH, then a second change in a few hours if it is still over. Add a conditioner; add air. If the tank is cycling with fish in, it is time to rehome or move them. |
The words are the calculator's own verdict states, so the table and the result can't disagree. None of this is a diagnosis: if fish are gasping at the surface, flashing or lying on the bottom with ammonia at zero, look elsewhere — nitrite, oxygen, temperature — and the quarantine guide says when to ask someone.
Common questions
My kit always reads a little ammonia and the fish are fine — why? Three usual reasons. Chloramine-treated tap water leaves ammonia behind once the conditioner has broken it apart; a Nessler-type kit reads the ammonia a conditioner has bound as if it were still free; and a soft, acidic tank genuinely can carry a fraction of a ppm with almost none of it in the toxic form — run the numbers at your pH and see. Below pH 6.5 the filter also slows down, which is its own reason a low-pH tank shows a persistent trace.
So is ammonium safe? Safer, not safe. NH₄⁺ is far less toxic at the concentrations a kit reads, but it is a reservoir: raise the pH and it becomes NH₃ on the spot. A tank at pH 6.2 holding 2 ppm is one water change with alkaline tap water away from an emergency. Treat the total as the thing to remove and the pH as the thing not to move while you do it.
Which tank types are most exposed? Anything kept alkaline on purpose: Rift lake cichlids (7.8–8.6), livebearers, goldfish in hard water, and every marine tank. Their budget is a fraction of a ppm, which is why an ammonia reading in those tanks is a drop-everything event and the same number in a blackwater tank is something to fix this week.