How much ammonia for a fishless cycle?
Enough to reach 2–4 ppm in your tank, then enough to put it back each time the bacteria eat it. With the usual 10% janitorial bottle, that first dose is smaller than people expect:
| Tank | Water | 10% ammonia | In drops | Or dry NH₄Cl |
|---|---|---|---|---|
| 10 gal | 38 L | 0.8 mL | 16 drops | 0.24 g |
| 20 gal | 76 L | 1.6 mL | 32 drops | 0.48 g |
| 29 gal | 110 L | 2.3 mL | 46 drops | 0.69 g |
| 40 gal | 151 L | 3.2 mL | 63 drops | 0.95 g |
| 55 gal | 208 L | 4.3 mL | 87 drops | 1.31 g |
| 75 gal | 284 L | 5.9 mL | 119 drops | 1.78 g |
| 125 gal | 473 L | 9.9 mL | 198 drops | 2.97 g |
To 2 ppm as NH₃, computed with this site's engine from the density of 10% aqueous ammonia (95.8 mg NH₃ per mL) and the molar mass of NH₄Cl. Drops at 0.05 mL each — count them for small tanks, use a syringe over about 20 gallons.
Percentages on chemical labels are by weight, and ammonia solution is lighter than water — 10% ammonia has a density of about 0.9575 g/mL, so a milliliter carries 95.75 mg of NH₃, not 100. Assuming otherwise overdoses you by 4.4%. Small, but it is free to get right and it compounds every time you re-dose.
Which ammonia should I buy?
Three things work, and the expensive one is not better. What matters is only that it is ammonia and nothing else:
- Household or janitorial ammonia, 10%. A few dollars for a bottle that will cycle a tank many times over. It must be plain — no detergent, no perfume, no color, nothing labeled "sudsy" or "cloudy". Surfactants foam in the tank, coat surfaces, and are miserable to remove. Shake the bottle: real ammonia settles at once.
- Dry ammonium chloride (NH₄Cl). The most precise option, because a solid does not have a strength to guess at and does not weaken on the shelf. It needs a scale that reads to 0.01 g — for a 20 gallon the dose is 0.48 g, which no kitchen spoon will measure honestly.
- Branded "cycling ammonia". Usually ammonium chloride in water, sold at a large markup with a dosing chart on the label. It works, and it removes the guesswork about what is in the bottle. If the label states a dose, use it — the maker knows their own concentration better than you can infer it.
What does not work: fish food left to rot (it makes ammonia, but slowly, unpredictably, and it also makes phosphate and a bacterial bloom), a raw shrimp (same, plus a smell), or "cycling in a bottle" products used on their own without an ammonia source. There is nothing wrong with bottled bacteria alongside a real dose — but bacteria still need feeding, and an unfed colony does not grow.
"Dose to 2 ppm" — 2 ppm of what?
This is the trap nobody warns about. An ammonia figure can be quoted as NH₃ (the whole molecule) or as nitrogen (NH₃-N, counting only the nitrogen atom). They differ by a factor of 1.216, and cycling instructions almost never say which they mean.
In practice: if you dose to 2 ppm meaning nitrogen, a kit reporting as NH₃ will read 2.43 ppm. Go the other way and you are under-dosing by nearly a fifth. On its own that is survivable — but it eats the margin between a normal dose and the 5 ppm mark where ammonia starts inhibiting the very bacteria you are trying to grow. Somebody aiming at "4 ppm" in the wrong basis lands near 4.9.
The fix is dull and reliable: use one kit and one basis for the whole cycle, and don't mix a dosing instruction from one source with a reading from another. The toggle above the target field switches this calculator between the two, and the "show the math" panel prints whichever you picked.
Why cycles stall: the alkalinity the ammonia eats
Nitrification is an acid-producing reaction. Every mole of ammonia oxidized to nitrate releases two moles of hydrogen ions, and those consume your carbonate hardness:
NH₄⁺ + 2 O₂ → NO₃⁻ + 2 H⁺ + H₂O
Two equivalents of alkalinity per nitrogen works out at 7.15 mg of CaCO₃ per mg of ammonia nitrogen — which for a 2 ppm dose is 0.66 dKH. One dose is nothing. A cycle is not one dose.
| Starting KH | 2 ppm doses before it stalls | What that means |
|---|---|---|
| 2 dKH | no headroom | The first dose already takes it under 3 dKH. Buffer before you start. |
| 3 dKH | no headroom | The first dose already takes it under 3 dKH. Buffer before you start. |
| 4 dKH | 1 | Runs out early — this is the classic nitrite stall waiting to happen. |
| 6 dKH | 4 | Runs out early — this is the classic nitrite stall waiting to happen. |
| 8 dKH | 7 | Close. Likely to need topping up once near the end. |
| 12 dKH | 13 | Enough buffer to ride out a normal cycle. |
Doses until KH reaches 3 dKH, below which nitrification slows sharply. Derived from the equation above; assumes nothing is buffering the tank back up.
This is the arithmetic behind the most common cycling complaint in the hobby — "my nitrite has been stuck for three weeks". A soft-water tank starting at 4 dKH has about 1 dose of road in front of it. After that the pH slides, the bacteria slow down, and nothing moves no matter how patiently you wait. The fix is a water change or a little baking soda to put the KH back — not more ammonia, which makes it worse.
A KH test is the cheapest insurance in a fishless cycle and almost nobody runs one. If you are on soft or RO-remineralized water, check it weekly through the cycle and keep it above about 4 dKH. Hard tap water usually carries enough buffer to ride the whole thing out without you thinking about it — which is exactly why the advice you read online may not have mentioned it.
The nitrate you're manufacturing
Every milligram of ammonia nitrogen ends up as 4.43 mg of nitrate, and nothing removes it for you. One 2 ppm dose makes about 7.3 ppm of nitrate; a cycle of 8 doses leaves you sitting on roughly 58 ppm.
That is not a problem — it is the proof the cycle worked — but it does mean the tank is not ready the moment the test passes. Do a large water change to bring nitrate down before livestock goes in, and then stock promptly, because a cycled filter with nothing to eat starves back down within days. The water change calculator will size that final change from your actual reading.
I overdosed — now what?
Common, and fixable. If you have overshot past about 5 ppm, the answer is a water change, because there is nothing you can add that removes ammonia without also removing it as food. Dilution is exactly the arithmetic the water change calculator does: to come down from 8 ppm to 4 ppm you replace half the water, since your source water contains none.
Don't dechlorinate the tank with an ammonia-binding conditioner to fix it. Products of that type convert ammonia to a form the test kit may still read but the bacteria consume more slowly, which leaves you unable to tell what is happening — the worst possible state during a cycle. Change water instead, and keep the conditioner for the new water going in.
A useful sanity check before you dose anything: the first dose in the table above is 1.6 mL for a 20 gallon. If your bottle's instructions or an online figure suggest something several times that, one of you has the strength or the basis wrong — work it out before it goes in the water.
When do I re-dose?
Whenever ammonia hits zero, and not on a schedule. Dosing on a calendar is how people end up above 5 ppm and wonder why progress stopped. Early on that will be every few days; by the end the tank will clear a full dose overnight, which is the signal you are close.
If you want that as a rule rather than a judgment call, the step-by-step guide has a table of every reading you can get in the morning and what to do about it that day — plus a printable log to keep the readings in.
Knowing when it is finished is a different question from how much to dose, and it has one unambiguous answer — the 24-hour test. The nitrogen cycle guide sets it out, along with what each stage looks like on a test kit and what to do if the numbers stop moving.