How to Cycle a New Fish Tank (Before a Single Fish Goes In)
By Yechiel Kuperman, marine biologist · 49 years keeping fish
Quick answer
Cycling means growing the bacteria that turn fish waste into safer compounds before any fish arrive, because a brand-new filter has almost none.
The kindest way is fishless: add a measured dose of ammonia to the running tank, test every few days, and top it up as the bacteria eat it.
It usually takes four to eight weeks, and it is finished when the tank turns 2 mg/L (2 ppm) of ammonia into zero ammonia and zero nitrite within 24 hours.
The numbers
| What | Metric | US | Why |
|---|---|---|---|
| Ammonia dose | 2 mg/L total ammonia | 2 ppm total ammonia | Top back up to 2 when it drops. A bigger dose stalls the cycle |
| Water temperature while cycling | 24–28 °C | 75–82 °F | The bacteria grow faster warm. A cold tank cycles slowly |
| pH while cycling | about 7 to 8 | same scale everywhere | The cycle itself pushes pH down, and acidic water slows it |
| Typical time | 4–8 weeks | 4–8 weeks | UF/IFAS gives six to eight weeks for a new biofilter |
| The finish line | 2 mg/L gone in 24 hours | 2 ppm gone in 24 hours | Zero ammonia AND zero nitrite the day after a full dose |
| Nitrate before the first fish | under 40 mg/L | under 40 ppm | One big water change on the last day brings it down |
| Fish-in cycle: change water when | ammonia or nitrite above 0.25 mg/L | above 0.25 ppm | Test daily and change 25–50% as often as it takes |
Do this — step by step
- Fill the tank, dechlorinate the water, and run the filter day and night from now on.The bacteria live where water flows
- Set the heater to the temperature your future fish will need.Warmer water cycles faster
- Dose pure ammonia to 2 mg/L (2 ppm).Label must say ammonia only: no scent, no soap
- Test ammonia, nitrite and pH every two or three days, and write every number down.
- Whenever ammonia falls, dose it back up to 2 mg/L.This is the bacteria's food
- Wait for nitrite to rise and then fall to zero.The slow half. Keep dosing
- When a full 2 mg/L dose reads zero ammonia and zero nitrite 24 hours later, you are cycled.
- Change a big share of the water, then add a small first group of fish within a day or two.Unfed bacteria start to die back
- Never rinse the filter sponge under the tap, now or ever.Old tank water only
What is going on inside a tank while it cycles?
Professor Nitra's line in Yechiel's printed guide is the whole mechanism in two sentences. He converts toxic ammonia into nitrite. His colleagues convert nitrite into nitrate. Cycling is the weeks it takes to grow enough of both teams.
The first team, the ammonia oxidisers, lives as a thin film on the filter sponge, the gravel and the glass. They get their energy by pulling electrons off ammonia, which is a meagre living. So they grow slowly, far slower than the bacteria that turn a dropped pellet into mush overnight.
The second team eats nitrite. In freshwater aquarium filters, DNA sampling found these are mostly Nitrospira, not the Nitrobacter printed on older charts (Hovanec et al., 1998). Since 2015 we have also known that some Nitrospira can do both jobs alone.
The second team can't grow until the first has made nitrite for it to eat, and that gives every fishless cycle log the same shape. Ammonia starts falling. Nitrite climbs, peaks and then drops. Nitrate climbs last, and keeps climbing.
The nitrite phase is where most people give up. It is usually the longer half, and a column of identical nitrite readings for a week or more looks exactly like failure. Keep dosing and keep testing. It moves.
Nitrification also eats alkalinity, the water's buffer against acid. As the bacteria work, the pH creeps down, and in acidic water they slow further. UF/IFAS describes the extreme version as old tank syndrome, below about 20 mg/L alkalinity. A water change restores the buffer.
Dose matters too. Very high ammonia inhibits the nitrite-eating bacteria, a finding from wastewater work (Anthonisen et al., 1976). That is why 2 mg/L works better than pouring in more: a moderate dose feeds the colony without holding the second team back.
Why go fishless at all? Because the alternative puts fish into water that is guaranteed to fill with the two things that hurt them most. Ammonia burns gills. Nitrite turns haemoglobin into methaemoglobin, which can't carry oxygen. A fish-in cycle is survivable with daily testing, but it is hard on the fish.
Bottled bacteria and borrowed filter media from a healthy, established tank can both shorten the wait. Neither replaces the 24-hour test at the end, because that test is the only thing that proves the colony is big enough.
When the test passes, the colony is sized to about a 2 mg/L daily load. That is why the first fish should arrive soon, in a small group. The bacteria grow to match what they are fed, and they need a week or two to catch up with every new batch.
By the way, out in the ocean…
Killer whale (orca). A powerful apex hunter that is actually the largest dolphin — living in close family pods, each with its own dialect. Orca pods pass what they know down the generations. Filters can too: media from a healthy tank gives a new one a head start.
Jellyfish. Drifting through the sea with no brain and no heart — and some kinds glow with their own light. Nitrate, the end product of your cycle, is plant food. In the sea, extra nitrogen feeds the plankton that jellyfish blooms can follow.
Professor Nitra’s Logbook Card
Nitrifying bacteria · resident of the filter housing
I would like it noted that I cannot be bought in a shop, rushed by a deadline, or persuaded by a child pressing her nose to the glass. I grow. Slowly. On a sponge, in the dark, with water pouring over me.
Feed me ammonia and I will make nitrite. My colleagues will make nitrate out of that, once they get going, which takes longer than anyone would like, including them.
When the 24-hour test passes, bring the fish. Not all the fish. A few. And the next person who rinses my sponge under the tap will hear from me very briefly, because that is roughly how long I will survive it.
What goes wrong during a first cycle
Adding fish on day one
A tank filled yesterday has water, a filter and almost none of the bacteria that make water safe. The first fish start producing ammonia at once, and nothing is there to eat it.
Ammonia climbs, nitrite follows, and fish sicken in the first weeks. It is often called new tank syndrome, and cycling first is the whole cure.
Using household ammonia with extras in it
Many cleaning ammonias contain scent or surfactants that have no business in a fish tank. Shake the bottle. If it foams, leave it on the shelf.
The label should list ammonia and water, nothing else. Aquarium shops sell bottles made for exactly this job.
Deciding the cycle has stalled and starting again
Nitrite that sits high for a week feels stuck. Usually it is the slow second team doing its slow thing. Restarting throws away every bacterium you grew.
Check the pH instead. If it has slid well below 7, a water change lifts it, and so does easing back on the ammonia dose.
Cleaning the filter halfway through
A cycling filter is barely dirty, and the film on the sponge is the entire project. Rinsing it under the tap wipes out weeks of growth in seconds.
Passing the test, then waiting a month for the fish
Bacteria with nothing to eat die back. If the fish are delayed, keep dosing ammonia every day or two until they arrive, then stock lightly.
Questions we get asked
How long does it take to cycle a fish tank?
Usually four to eight weeks. UF/IFAS gives six to eight weeks for a new biofilter, and a warm tank with a steady ammonia supply sits at the faster end.
Seeding with media from an established tank can cut that a lot. The 24-hour test tells you when you are done, not the calendar.
How do I know when my tank is cycled?
Dose ammonia to 2 mg/L (2 ppm) and test 24 hours later. If ammonia and nitrite both read zero and nitrate has risen, the bacteria can handle that load and the tank is cycled.
Anything above zero means keep going, dosing and testing as before.
Can I cycle a tank with fish in it?
Yes, but it is hard on them. Use only a few hardy fish, feed lightly, and test ammonia and nitrite every day.
Change 25 to 50 percent of the water whenever either reads above about 0.25 mg/L. A fishless cycle avoids all of that, which is why we recommend it.
Do bottled bacteria really work?
Some do. Products with live nitrifying bacteria can shorten a cycle, and a handful of used filter media from a healthy tank does the same job for free.
Quality varies, and a bottle proves nothing on its own. Run the 24-hour ammonia test before you add fish, whatever the label says.
Why is my nitrite so high and not going down?
The nitrite phase is usually the slowest part, so patience comes first. Check pH next: if it has slid below about 6.5, the bacteria slow down and a water change will lift it.
Also check you are not overdosing ammonia, because very high levels hold the nitrite-eaters back.
Can I cycle a tank with fish food instead of ammonia?
You can. Rotting food releases ammonia, so a pinch a day will start a cycle. It is harder to control, though, because you cannot tell how much ammonia you are adding.
It also fouls the tank. Pure ammonia dosed to 2 mg/L gives you a clean, readable test log.
Does a water change reset the cycle?
No. The bacteria live as a film on the filter, gravel and glass, so changing water barely touches them.
What does hurt them is chlorinated water and tap-rinsed sponges. Always dechlorinate, match the temperature, and clean filter media only in old tank water.
What is new tank syndrome?
It is what happens when fish go into a tank before the filter bacteria exist. Ammonia builds up with nothing to remove it, then nitrite follows, and fish sicken or die in the first weeks.
It is one of the commonest ways a first aquarium fails, and cycling first prevents it.
Sources
- Aquarium Basics, Part 5Kuperman, Y. Aquarium Basics, Part 5 of 9: The Filter, a System of Life. Yechiel's Fishery Farm care guide (PDF). View source
- UF/IFAS FA031Francis-Floyd, R., Watson, C., Petty, D. & Pouder, D. B. Ammonia in Aquatic Systems. University of Florida IFAS Extension, FA16/FA031. View source
- Hovanec et al. 1998Hovanec, T. A., Taylor, L. T., Blakis, A. & DeLong, E. F. (1998). Nitrospira-like bacteria associated with nitrite oxidation in freshwater aquaria. Applied and Environmental Microbiology 64(1), 258–264. View source
- Daims et al. 2015Daims, H., Lebedeva, E. V., Pjevac, P., Han, P., Herbold, C., Albertsen, M., et al. (2015). Complete nitrification by Nitrospira bacteria. Nature 528, 504–509. View sourceNote: Shows a single Nitrospira can take ammonia all the way to nitrate. It does not tell us how common that is inside any particular aquarium filter.
- Anthonisen et al. 1976Anthonisen, A. C., Loehr, R. C., Prakasam, T. B. S. & Srinath, E. G. (1976). Inhibition of nitrification by ammonia and nitrous acid. Journal of the Water Pollution Control Federation 48(5), 835–852.Note: A wastewater study. It establishes that free ammonia inhibits nitrite oxidisers. The exact thresholds in a home aquarium will differ.
- SRAC 462Durborow, R. M., Crosby, D. M. & Brunson, M. W. (1997). Nitrite in Fish Ponds. Southern Regional Aquaculture Center Publication No. 462.Note: The source for the methaemoglobin mechanism. A pond document: use the mechanism, not the dose rates.
- DrTim's AquaticsDrTim's Aquatics. Fishless Cycling. Guidance on dosing to 2 ppm and the 24-hour zero-ammonia, zero-nitrite completion test. View sourceNote: Published by a company that sells bottled nitrifying bacteria (founded by the lead author of Hovanec et al.). Used here for the method and the completion test only, not for any product claim.
Want the original printed guide?
Download the PDFAbout the author
Yechiel Kuperman
Marine Biologist & Aquaculture Mentor
Yechiel has kept fish since he was four — 49 years with tanks — and 30+ years running farms. He mentors fish farmers internationally.
He writes the Yechiel's Fishery Farm books, in which the biology is real and the fish have opinions.
Last checked: