Fishless cycling, step by step

Every step-by-step cycling guide on the internet gives you a calendar — ammonia falls around day nine, nitrite peaks around day sixteen, fish around day thirty. Those numbers are real, in the sense that they happened to somebody. They are also useless to you, because the same tank cycles in 2.5 weeks or 7.9 depending on things most of those guides never mention.

So this page is a loop with decision rules instead of a timetable, and then the arithmetic of what actually sets the pace. If you want the biology first, read the nitrogen cycle explained and come back.

What you need before day one

Short list, and only one item on it is worth agonizing over:

• A liquid test kit that does ammonia, nitrite, nitrate, pH and KH. This is the whole project. A cycle you cannot measure is a cycle you cannot finish, and test strips are not accurate enough at the low end where the answer lives — "not quite zero" and "zero" look identical on a strip and mean completely different things.
• An ammonia source. Plain household ammonia or dry ammonium chloride; the cycling calculator covers how to tell a usable bottle from a useless one and converts either into a dose for your tank.
• Dechlorinator, for the fill and for every top-off after it. Chlorine and chloramine are designed to kill exactly what you are growing.
• A heater, running. See how long it actually takes — this is the single biggest lever you have and most people leave it at 78 °F out of habit.
• Somewhere to write numbers down. There is a printable log below. Six weeks of readings held in your head become "I think it went down a bit?", which is not a measurement.

Day one: set it up, then feed it

1. Build the tank completely. Substrate, hardscape, filter, heater, lid — everything, in its final position. The bacteria colonize surfaces, so anything you add later is bare, and anything you take away leaves with its share of the colony.
2. Fill and dechlorinate. Treat the full volume, not the bucket you happen to be holding.
3. Turn everything on and leave it on. Filter running continuously — no overnight breaks, no "resting" the pump. These bacteria are aerobic and the filter is where most of them will live.
4. Set the heater warm. 82–86 °F (28–30 °C) if it will hold it. Nothing in the tank objects.
5. Test your tap water before you dose anything — pH, KH, and nitrate. The KH reading is the one that decides whether this goes smoothly, and the nitrate reading is what you will be comparing against at the end.
6. Dose ammonia to 2 ppm and write down the date, the dose, and the reading you got an hour later. That hour matters: it is your proof the dose landed where you meant it to, and it calibrates every re-dose afterwards.

Then stop. The single most common day-two mistake is adding more ammonia because "nothing is happening". Nothing is supposed to be happening yet.

Then it's a loop, not a calendar

From here you do the same three things every morning — test ammonia and nitrite, write them down, decide — and the decision is entirely determined by the two numbers. The tank tells you what day it is.

Ammonia readings assume one kit and one reporting basis throughout — mixing an "as N" instruction with an "as NH₃" kit is a 22 % error, which the calculator explains.

Two rules carry most of the weight here. Re-dose at zero, never on a schedule — dosing every day is how a tank ends up at 8 ppm with a stalled cycle and an owner who thinks they are being diligent. And test nitrite before you celebrate: ammonia hitting zero in a day is the halfway mark, not the finish, and the second colony routinely takes twice as long to establish as the first.

Track it on one sheet of paper

Six weeks of daily readings is a data set, and the shape of it is what tells you whether you are progressing or stuck. Print this, tape it inside the cabinet door, fill in a row a day.

How long will it actually take?

4 weeks is the honest ballpark for a warm, buffered tank started from nothing. But the pace is set by how fast a bacterial population doubles, and that is a strong function of temperature — strong enough that it is the difference between a month and a season.

Wastewater engineering has had this number since 1964, which is why it is odd that the aquarium hobby quotes it as "warmer is better". Nitrifier growth rate rises about 10 % per °C, so:

Computed from the nitrifier growth-rate equation used in activated-sludge design (Downing & Hopwood 1964, carried into the USEPA nitrogen-control design manuals): growth ∝ e^{0.098(T−15)}. Weeks are that ratio applied to the hobby's observed 4 weeks at 77 °F, at pH 7.5.

Read the ratios, not the decimal places. The point is the spread: 64 °F takes 3.2 times as long as 86 °F, same tank, same bottle, same everything else. A spare-room tank in an unheated house is not being slow because you did something wrong. It is being slow because it is cold.

Why an acid tank crawls, and then stops

The same equation has a pH term, and it is unforgiving. Below pH 7.2 the rate falls off in a straight line, and the line reaches zero at pH 6.0:

The pH term of the same design equation, [1 − 0.833(7.2 − pH)], at 77 °F. Real biofilms keep working a little below pH 6 — but only a little, and by then the cycle has stopped being a waiting problem and started being a chemistry one.

This is why the two levers compound so badly. A tank at 68 °F whose pH has drifted to 6.6 is looking at roughly 13 weeks — the three-month cycle people post about in despair, and neither half of it is bad luck.

And the pH drop is usually self-inflicted, in the sense that nitrification causes it: the reaction produces acid and eats your KH as it goes. Once KH is gone the pH falls, and the falling pH stops the bacteria that were consuming the KH. The cycling calculator works out exactly how many doses your starting KH can absorb before that happens — for a lot of soft-water tanks the answer is fewer than the cycle needs, which is worth knowing on day one rather than week five.

The one-week version: seeding

Everything above is about growing a colony. The fast route is to not grow one:

• Established filter media from a healthy tank — a squeezed sponge, a bag of ceramic media, a handful of mature substrate. This is transplanting a working population rather than culturing one, and it can take a cycle from weeks to days. Ask at a local club or a good independent store; most keepers will hand you a sponge for nothing.
• Bottled bacteria, honestly: some products contain live cultures of the right organisms and genuinely compress a cycle, and several have been tested independently and shown to work. Others contain the wrong species, or the right ones cooked in a warehouse. Treat a bottle as something that might help, keep dosing ammonia and keep testing, and never stock on the strength of a label that says 24 hours.

Whatever you seed with, the finish test does not change. A seeded tank still has to take a full dose to zero and zero in a day, and it is the only evidence that counts.

The last two steps, which everyone rushes

You have had a 24-hour clearance and the tank is cycled. Two things left, and doing them in the wrong order or on the wrong day undoes the whole six weeks.

1. A big water change. Eight doses of 2 ppm manufacture around 58 ppm of nitrate, because nothing removes it for you. That is the proof the cycle worked and it is far too high to stock into. The water change calculator sizes it from your actual reading and your tap's own nitrate.
2. Fish within a day or two. The colony is sized to the amount of ammonia it has been fed. Stop feeding it and it starves back down within days — and then the fish you add a fortnight later produce more waste than the survivors can process, which is a mini-cycle in a stocked tank. If the fish are delayed, keep dosing ammonia until the day before they arrive.

Stock in stages regardless. A cycle proved at 2 ppm a day supports roughly 2 ppm a day of livestock waste, and a full stocking on day one asks for more than that. How many fish that comes to is a separate question with a worse reputation — "one inch per gallon" is wrong, and the nitrate reading you are already taking answers it better.

Mistakes that add weeks

Not the biology — that list lives on the nitrogen cycle page. These are the procedural ones, in rough order of how much time they cost:

• Dosing on a calendar. Ammonia goes in when the test says zero, and not before. Daily dosing is the fastest way to an inhibited, stalled tank.
• Running it cold. Worth up to 3.2× the wait, and it is free to fix.
• Never testing KH. The most common cause of the three-week nitrite plateau, and the one test nobody runs until they are told to.
• Adding an ammonia-binding conditioner to fix a high reading. It converts ammonia to a form your kit may still read but the bacteria consume more slowly — so you lose the ability to interpret your own tests, which is the only instrument you have. Change water instead.
• Rinsing the filter media, or replacing a cartridge, mid-cycle. You are throwing away the thing you are growing.
• Cycling, then waiting. A finished cycle is perishable. Keep feeding it or use it.

What you'll be told, and what's true

• "Add ammonia every day to keep it fed." This is the single most common way a fishless cycle goes wrong. Before the bacteria exist, nothing is consuming yesterday's dose, so daily dosing stacks — three days in you are past 5 ppm and actively suppressing the colony. Dose at zero, and only at zero.
• "Dose to 4 ppm to build a bigger colony." Half true, and the half that is wrong is expensive. A bigger daily feed does support a bigger colony, which is why heavily-stocked plans cycle at the top of the range. But the effect reverses above about 5 ppm, and it costs proportionally more KH — so "more to be safe" buys a stall, not a margin.
• "Cycling takes six weeks. That's just how it is." Six weeks is what an unheated, unbuffered tank takes. The same tank at 86 °F with intact KH is on the order of 2.5. The duration is mostly a decision you made without realizing it was one.
• "Bottled bacteria cycle a tank in 24 hours." Some products really do compress a cycle dramatically, and the good ones have independent testing behind them. But "24 hours" is a marketing claim, not a test result, and the only way to know is the same 24-hour clearance test you would run anyway. Use the bottle; do not skip the test.

Common questions

Do I need to do water changes during a fishless cycle?

Normally no — there are no fish to protect, and changing water only removes the ammonia you are paying to add. There are two exceptions. If ammonia has gone above about 5 ppm you are inhibiting the bacteria and a change fixes it. If KH has fallen below about 3 dKH the cycle is about to stall, and a change restores the buffer. Both are corrections, not maintenance.

Can I fishless cycle a planted tank the same way?

Yes, with one caveat that catches people out: plants take up ammonia directly and prefer it to nitrate, so your ammonia reading falls partly because the plants ate it, not only because bacteria did. That makes the ammonia test optimistic. Judge a planted tank on nitrite instead — nitrite appearing and then clearing is bacterial work that plants do not do for you. Dosing to 2 ppm is safe for almost all plants.

Should I run the heater hot for the whole cycle?

It is the cheapest accelerator you have, and there are no fish present to mind. Run it around 86 °F (30 °C) if your heater will hold it, then bring it back to the tank's normal temperature over a day or two before stocking. Do not leave it hot and drop it the morning the fish arrive — the colony is sized to the rate it has been working at, and a sudden 15-degree drop slows it just as you start feeding it real waste.

What if I have to go away for a week in the middle?

Dose to the top of the range (4 ppm) on the way out and leave it. You will come back to a tank that has processed the dose and then sat idle for a few days, which costs a little momentum but nothing that does not recover in a day. Do not dose extra "to cover the week" — anything above about 5 ppm works against you, so the week idle is cheaper than the overdose.