Key takeaways
- Dose to 2 ppm ammonia rather than the older 4 or 5 ppm, because high concentrations inhibit the nitrite oxidisers.
- Seeding a third of the media from an established filter cuts a five week cycle to between five and fourteen days.
- The nitrite plateau can sit at the top of the test card for ten days and needs no intervention at all.
- A cycle is finished only when a 2 ppm ammonia dose clears to zero ammonia and zero nitrite within 24 hours.
- Nitrification stalls below pH 6.5 and below 4 dKH, so soft water tanks need bicarbonate to keep the buffer up.
A bottle of unscented household ammonia costs around $4 and will bring a filter to full working capacity in three to five weeks with no fish in the tank at all. Hardly anyone buys one.
That is why the standard first month in this hobby involves daily water changes, a test kit that keeps reading 1 ppm, and two or three fish that do not make it to week four.
What you are growing, and why it takes weeks
Two groups of bacteria have to establish in sequence. The first oxidises ammonia into nitrite, and the second oxidises nitrite into nitrate.
The second group cannot start colonising until the first group has produced enough nitrite to feed it. That dependency is why the process has two distinct humps rather than one, and why it cannot be rushed past a certain point.
Both groups are slow growers by bacterial standards, doubling in something like 15 to 24 hours under good conditions rather than the 20 minutes a gut bacterium manages. They are also aerobic, which makes surface agitation part of the setup rather than an afterthought.
Why doing it with fish is the bad version
Fish-in cycling works, in the sense that the filter eventually matures. The fish pay for it in gill damage and immune suppression while ammonia sits at levels no test kit is meant to see on a running tank.
Ammonia toxicity climbs sharply with pH and temperature, because more of the total shifts into the un-ionised form that crosses gill membranes. At pH 8.2 and 27 C (81 F) a reading of 1 ppm total ammonia is far more dangerous than the same reading at pH 6.8. The EPA ammonia criteria pages set out that relationship in detail.
You also cannot stop it going wrong. Water changes to protect the fish remove the ammonia the bacteria need, which stretches the process out, which extends the exposure.
What to have ready before you start
- A liquid test kit reading ammonia, nitrite, nitrate and pH. Strips are not accurate enough for this job.
- An ammonia source: unscented household ammonia, or a dedicated product such as Dr Tim’s Ammonium Chloride or Fritz Fishless Fuel.
- A heater holding 26 to 28 C (79 to 82 F), which is the fastest safe range.
- Dechlorinator, and a KH test if your tap water is soft.
- Seed media from an established filter, if you can get any.
Shake the ammonia bottle before buying if you can. A head of foam that sits there for more than a few seconds means surfactants, and surfactants will foul the tank. The ingredient list should say ammonia and water, nothing else.
Getting the dose right
Aim for 2 ppm total ammonia. Older guides recommend 4 or 5 ppm, and there is reasonable evidence that concentrations that high inhibit the nitrite oxidisers rather than feeding them, which stalls the second half of the process.
Household ammonia is sold at anything from 5 to 10 percent, so no single dosing figure works across brands. Add roughly 1 ml per 40 litres, wait an hour with the filter running, test, then scale from what you measured. Write the working dose on the bottle with a marker.
Dose to 2 ppm on day one and then leave it alone. Do not top up daily at the start, because nothing is consuming it yet and you will simply climb past the useful range.
Do not dose an ammonia binding conditioner during a fishless cycle. Products of that type interfere with the salicylate and Nessler chemistries most hobby test kits use, and you will spend a fortnight chasing readings that are not real.
The curve, week by week
| Stage | Typical timing | Ammonia | Nitrite | What to do |
|---|---|---|---|---|
| Dosed and waiting | Days 1 to 8 | 2 ppm, flat | 0 | Nothing. Test every second day. |
| First oxidisers working | Days 8 to 18 | Falling | Rising | Redose to 2 ppm when ammonia hits 0. |
| Nitrite peak | Days 14 to 28 | Cleared in 24 h | 2 to 5 ppm or off scale | Hold. This stage feels stuck and is not. |
| Nitrite falling | Days 21 to 35 | 0 | Dropping fast | Keep redosing ammonia daily. |
| Cycled | Days 21 to 45 | 0 in 24 h | 0 in 24 h | Large water change, then stock. |
The nitrite plateau is where most people give up. It can sit at the top of the test card for ten days or more while looking exactly like a failure, and the correct response is to change nothing.
Once ammonia clears within 24 hours, start redosing to 2 ppm every day. That keeps the first group fed while the second group catches up.
Seeding from an established filter
Nothing else comes close for speed. A third of the ceramic media from a mature canister, or the squeezings of a mature sponge, cuts a five week job to somewhere between five and fourteen days.
Move it wet and move it fast. Biological media in a sealed bag goes anaerobic within a few hours, so carry it submerged in tank water, in an open container, and have it in the new filter inside an hour. A donor sponge filter kept running in another tank is the tidiest version of this, because you can move the whole unit.
Take no more than a third from the donor filter and the source tank will not notice. Take more and you have two half-cycled systems instead of one working one.
Refrigerated bottled bacteria such as Fritz TurboStart 700 or Dr Tim’s One and Only contain the right species and can take a cycle down to 7 to 14 days. Shelf-stable bottles that have sat in a warm shop for months are far less reliable. Check for a cold chain and a date.
Conditions that stall a cycle
Four things account for nearly every stalled fishless cycle, and three of them are chemistry rather than bacteria.
- pH below 6.5. Nitrification slows markedly and can halt near pH 6.0. The process itself consumes alkalinity, so soft water tanks drift downward as they go.
- KH below about 4 dKH. Add half a teaspoon of bicarbonate of soda per 100 litres to rebuild buffer if pH is sliding.
- Temperature under 20 C (68 F). Colony growth roughly halves for every several degrees you drop below the mid twenties.
- Ammonia pushed past 5 ppm. Back it off with a 50 percent water change and resume at 2 ppm.
Low oxygen is the fourth and the least obvious. A filter outlet held below the water line with no surface movement starves an aerobic process, so raise the outlet or add an airstone for the duration.
A 120 litre tank cycled in nine days
A customer setting up a 120 litre planted tank already had a mature 60 litre running in the same house. We took the middle basket of ceramic rings out of the small tank’s canister, roughly 700 ml, and dropped it straight into the new filter in a bucket of its own tank water.
Ammonia dosed to 2 ppm on the Saturday cleared in 30 hours. Nitrite showed on day three, peaked around 1.5 ppm on day five, and read zero on day eight.
Day nine, a 2 ppm dose cleared completely with no nitrite by the following evening. He did a 70 percent water change to drop nitrate from about 70 ppm and stocked eight fish that weekend. The donor tank never registered a reading, because porous biological media holds far more colony than a 60 litre tank needs.
The test that says it is finished
One test, and it is not negotiable. Dose to 2 ppm ammonia, wait 24 hours, and read both ammonia and nitrite. Both must be zero.
Ammonia at zero with nitrite at 0.25 ppm is not done, it is close. Repeat the dose daily until both read zero on the same test, which usually takes another two to four days.
Nitrate at the end will be somewhere between 40 and 160 ppm, which is expected and harmless to nobody because nothing is in there. Change 70 to 90 percent of the water, match the temperature, dechlorinate, and stock within a day or two so the colony does not go hungry.
Stocking without undoing the work
A filter cycled at 2 ppm of ammonia a day supports a moderate stocking level, not a full one. Add half your planned fish first, test daily for a week, and add the rest once you have seen three consecutive zeroes.
Feed lightly for the first fortnight, once a day and less than you think. Then leave the media alone: the most common way people undo a fresh cycle in month two is an enthusiastic first service, which is why the cleaning intervals for each media type matter more in a new tank than an old one.
Keep the media order sensible from the start as well, since the stack you build on day one decides how much of the colony survives the first year. The nitrogen chemistry behind all of this is covered plainly by the USGS water science school if you want the underlying reactions rather than the aquarium version.
Questions readers ask about this
Start with about 2.5 ml, wait an hour with the filter running, then test. Household ammonia is sold between 5 and 10 percent, so brands differ by a factor of two and no universal figure exists. Once you know the dose that reaches 2 ppm, write it on the bottle in marker.
Yes, and it is a good way to prepare a spare filter. Use 20 to 40 litres of dechlorinated water, a heater at 26 C (79 F), and dose ammonia the same way. The colony lives on the media, not in the water, so the finished filter transfers to any tank without losing capacity.
Only twice. Once if ammonia overshoots past 5 ppm or nitrite goes far off scale and the process visibly stalls, and once at the end to strip nitrate before stocking. Routine changes during the cycle just remove the food source and add days to the timeline.
Check pH first, because below 6.0 ammonia readings can fall for chemical reasons rather than biological ones. The other common cause is dosing an ammonia binding conditioner, which skews hobby test kits. Test your KH, add bicarbonate if it is under 4 dKH, and redose once pH is back above 7.
Roughly 24 to 48 hours out of water before losses become serious, and about a week running in water with no ammonia source at all. If a tank has to sit empty, keep the filter running and dose a small amount of ammonia, around 0.5 ppm, twice a week to keep the colony fed.
