Cycling a tank without poisoning fish.

The first tank I ever set up — at age twelve, with my mother's reluctant approval and a starter kit from a Sunday hardware run — killed three of its four goldfish in five days. The fourth survived, somehow, for two more weeks. We did everything wrong. Including the bit that nobody at the shop told us about: cycling.

Cycling is the process of establishing the colony of nitrifying bacteria that will convert your fish's waste into something less lethal. It takes three to five weeks. Without it, every fish you add is being slowly poisoned by its own ammonia. With it, your tank is stable from day one.

This is a long article because the nitrogen cycle is the foundation everything else sits on. If you've kept fish for years and never tested for ammonia, the only correction I have is: please start. If you're brand new and reading this before buying the fish — congratulations, you're already ahead of where I was at twelve.

The nitrogen cycle in three sentences

Fish excrete ammonia (NH₃ / NH₄⁺) through their gills and in their faeces. Two groups of bacteria — Nitrosomonas-like organisms first, then Nitrobacter-like and Nitrospira-like organisms — convert ammonia into nitrite (NO₂⁻) and then into nitrate (NO₃⁻). Nitrate is far less toxic and is removed by water changes (or, if you have a heavily planted tank, partly absorbed by the plants).

The "cycle" being referred to is that bacterial colony. A "cycled" tank is one where the bacteria are present in sufficient numbers to convert ammonia to nitrate as fast as your fish produce it.

Why ammonia and nitrite are both bad news

Both compounds damage fish gills, interfere with oxygen transport in the blood, and at higher concentrations cause death. Ammonia toxicity is also pH-dependent: at low pH (acidic water), more of the ammonia exists as NH₄⁺ which is less toxic; at high pH (alkaline water), more exists as NH₃ which is acutely poisonous. This is why a small ammonia reading in a hard, alkaline tank is a much bigger problem than the same reading in soft, acidic water.

The numbers I work to:

Reading	Acceptable for fish	Critical (act now)
Ammonia (NH₃/NH₄⁺)	0 ppm	Anything above 0.25 ppm
Nitrite (NO₂⁻)	0 ppm	Anything above 0.25 ppm
Nitrate (NO₃⁻)	Under 20–40 ppm	Above 80 ppm

"Under 20–40 ppm" reflects different schools of thought; planted-tank keepers tend to aim for the lower end, fish-only keepers accept up to 40. Above 80 ppm long-term causes chronic stress and stunting in many species.

Why "fishless" cycling

You can cycle a tank with fish, dosing carefully to keep ammonia under 0.25 ppm via constant water changes. People did it for decades. It's also functionally a slow form of animal cruelty, and we know enough now to do better.

Fishless cycling adds ammonia to the empty tank deliberately, lets the bacterial colony grow without any fish suffering in the process, and only introduces fish once the cycle is complete. It takes the same three to five weeks. The fish enjoy it considerably more.

What you need before starting

• The tank, fully set up, filter running. Substrate, hardscape, and any plants in place. Heater set to your target species' temperature (24–26°C is a reasonable default). Filter media properly seated.
• A liquid water test kit. The API Master Test Kit is the standard for good reason — strips are too imprecise to cycle by. You need accurate readings to ±0.25 ppm. Roughly £25 / $35.
• A source of ammonia. Either pure ammonia solution (Dr Tim's Ammonium Chloride is the cleanest; some hardware-store cleaning ammonia works if it's purely ammonia and water with no surfactants — read the label paranoid) or fish food (which decays into ammonia, slower and messier). I prefer the pure solution.
• A dechlorinator. Seachem Prime, Tetra AquaSafe, anything that neutralises chlorine and chloramine.
• Patience. Three to five weeks. It will feel longer.

The four-week protocol

Week 1 — Establishing the input

Once the tank is set up, water dechlorinated, filter running, and heater stable, dose ammonia to bring the tank to 2 ppm. With a typical 4% ammonium chloride solution, that's roughly 4 drops per US gallon (or ~1 ml per 10 litres — but check the label of your specific bottle, concentrations vary).

Test ammonia 24 hours after dosing. It should still read close to 2 ppm — there are no bacteria yet to consume it. Test nitrite and nitrate too. Both should read 0 ppm.

Re-dose ammonia every two or three days to keep the level around 2 ppm. The reasoning: the bacterial colony grows in proportion to the food supply. We're feeding the colony so it can build to a size capable of supporting a stocked tank.

By the end of week 1, you may see a tiny drop in ammonia. Most likely you won't yet. That's normal.

Week 2 — Ammonia falls, nitrite appears

Sometime between days 7 and 14 you'll start to see ammonia decreasing meaningfully between doses, and nitrite climbing. This is the first stage bacteria — the ammonia oxidisers — establishing themselves.

Continue dosing ammonia back up to 2 ppm whenever it drops below 1 ppm. Test daily. Record everything in a notebook (or your phone). The pattern you should see emerging:

• Ammonia: 2 ppm dropping to 0–1 ppm within 48 hours, faster as the week progresses
• Nitrite: rising, often to 2–5 ppm or higher
• Nitrate: starting to appear, slowly

Nitrite spikes are dramatic. Don't panic — they're expected and they mean the cycle is working. The first-stage bacteria are doing their job; the second-stage bacteria haven't caught up yet.

Week 3 — Nitrite falls, nitrate climbs

By day 14–20 you should see nitrite starting to fall and nitrate climbing more rapidly. The second-stage bacteria — the nitrite oxidisers — are now established. This is the slowest part of the cycle for many tanks; nitrite oxidisers reproduce more slowly than their ammonia-oxidising counterparts.

Keep dosing ammonia to 2 ppm. Test daily. The pattern you want to see:

• Ammonia: 2 ppm → 0 ppm within 24 hours
• Nitrite: starting to fall, perhaps 1–2 ppm at the time of testing
• Nitrate: climbing steadily, perhaps 20–40 ppm by end of week

Week 4 — The 24-hour confirmation

The cycle is complete when you can dose ammonia to 2 ppm and, 24 hours later, both ammonia and nitrite read 0 ppm. Nitrate will of course be elevated.

Test this twice over consecutive days to confirm. If both tests pass:

1. Do a large water change — 50% to 80% — to bring nitrate back down to under 20 ppm.
2. Do not dose ammonia again. Let the tank sit for 24 hours after the water change.
3. Add fish gradually. Start with a small group (4–6 small fish), test daily for the first week, and add the next group only after readings stay at 0 ammonia / 0 nitrite for a week.

Troubleshooting: when readings stall

The most common problems with fishless cycling, and what to do about them:

Ammonia isn't dropping after week 2

The first-stage bacteria haven't established. Possible causes: chlorine in your tap water (always dechlorinate), pH too low (below 6.5 slows the bacteria significantly — check your KH and consider raising it), temperature too low (raise to 26–28°C during cycling), or you've over-dosed ammonia (above 5 ppm inhibits the bacteria themselves).

Test pH. If it's below 6.5, do a 50% water change with conditioned tap water to nudge it up, then continue. Test ammonia after the water change — you'll likely need to re-dose to 2 ppm.

Nitrite is sky-high and not falling

This is the classic "stuck" point. Nitrite oxidisers are slower to establish, and high nitrite (above ~5 ppm) actually inhibits them. The fix: do a 50% water change to bring nitrite down, dose ammonia back to 2 ppm, and continue. Don't let nitrite climb above 5 ppm during cycling — it slows everything.

Cycle keeps "crashing" after fish are added

Usually one of three things: too many fish added at once (overloaded the bacterial colony), filter media replaced (the bacteria live on the media, not in the water — never replace all your media at once), or a treatment dosed that killed the bacteria (some antibiotics, copper-based parasiticides used at high doses, hydrogen peroxide flushes).

I added bottled bacteria. Did it work?

Sometimes. Tetra SafeStart and Dr Tim's One and Only contain live cultures of the right species and can shave roughly a week off cycling time when used correctly. Most other "instant cycle" products are between marginal and useless. If you use one, still test daily and confirm with the 24-hour zero readings before adding fish — bottled bacteria should accelerate cycling, not replace it.

Speeding things up: seeded media

The fastest legitimate way to cycle a new tank is to seed it with mature media from an established, healthy aquarium. A handful of ceramic rings or a piece of mature sponge from a friend's filter, dropped into your new filter at setup, will jump-start the bacterial colony and can reduce cycling time to as little as a week.

Caveats: the source tank must be visibly healthy (no ongoing disease, no recent treatments), and you'll still need to test daily through the cycle to confirm readings hit zero. Seeded cycling is faster but not skippable.

What this looks like in your test log

For my most recent setup — the 60-litre Apistogramma tank — here is the cycle log:

Day	Ammonia	Nitrite	Nitrate	Action
1	2.0	0	5	Dosed to 2 ppm. Tap nitrate baseline.
4	2.0	0	5	No change. Continue.
8	1.5	0.25	5	First sign of activity.
11	0.5	2.0	10	Stage 1 working. Re-dose to 2 ppm.
15	0.25	5.0	20	Nitrite climbing. 50% water change.
18	0	2.0	20	Stage 2 catching up. Re-dose.
22	0	0.25	40	Nearly there.
25	0	0	50	24h confirmation #1.
26	0	0	55	Confirmation #2. 70% water change.
27	0	0	15	First fish added (6 ember tetras).

26 days from start to first fish. Comfortably within the three-to-five-week window. The pair of Apistogramma cacatuoides went in three weeks later, after the embers had been showing zero readings the entire time.

The principle behind all of this

A cycled tank is a small, contained ecosystem. Once stable, it self-corrects within limits — small bioload variations get absorbed, the bacteria adjust to feeding patterns, and the tank runs quietly for years. An uncycled tank is a chemistry experiment with fish in it, and the experiment ends badly.

The discipline that pays off most in fishkeeping is the willingness to wait three weeks before adding a single fish. It's the difference between a hobby that's frustrating, expensive, and full of dead fish, and a hobby where the only mortality you see is old age. Use a calendar. Mark four weeks. Start now.

And while you wait — calculate your tank volume properly with our calculator, plan your stocking honestly with the stocking math article, and read up on the species you've decided on. The waiting time isn't dead time. It's research time.