Every new tank I have set up in the last decade, and there have been a lot of them, has gone through the same invisible process before a single fish went in. It is not glamorous. It involves staring at test tubes, waiting, and resisting the urge to buy fish for weeks on end. But cycling a tank properly is the single biggest difference between a smooth start in this hobby and the miserable, familiar story of new fish dying within a fortnight. This article is the version of the process I wish someone had handed me at the start: what is actually happening in the water, how to do it without fish, how long it genuinely takes, and the handful of mistakes that quietly wreck it.

What cycling actually means

Fish produce ammonia. It comes from their gills, their waste, and from any uneaten food or dead plant matter rotting in the tank. Ammonia is toxic at remarkably low concentrations. It burns gills, damages organs, and stresses fish to the point where diseases they would normally shrug off take hold.

In an established aquarium, that ammonia never gets the chance to build up, because two groups of bacteria deal with it. The first group consumes ammonia and converts it to nitrite, which is unfortunately also toxic. The second group consumes nitrite and converts it to nitrate, which is far less harmful and only becomes a problem at high concentrations over time.

Ammonia to nitrite to nitrate. That chain is the nitrogen cycle, and “cycling” a tank simply means growing enough of those two bacterial populations to process everything your fish produce, as fast as they produce it. The bacteria live mostly in your filter, coating the sponges and ceramic media in a thin film, with smaller populations on the substrate and every other surface in the tank.

A brand new tank has essentially none of these bacteria. Add fish on day one and their ammonia has nowhere to go. It accumulates, the fish suffer, and the shop gets blamed for selling weak stock. The stock was fine. The water was the problem.

Fishless cycling is the right default

The good news is that the bacteria do not need fish. They need ammonia, oxygen and a surface to live on, and you can supply the ammonia yourself. This is fishless cycling, and it should be the default method for every new tank, because nothing suffers while you learn.

Set the tank up completely first: filter running, heater set to somewhere around 24 to 26°C (the bacteria grow faster when warm), water dechlorinated. Then add an ammonia source. You have two realistic options.

The first is bottled ammonia. Look for plain ammonium chloride or household ammonia with no surfactants, perfumes or additives. Shake the bottle; if it foams and the foam lingers, it contains detergent and is useless to you. Several aquarium brands now sell ammonium chloride specifically for cycling, which removes the guesswork. Bottled ammonia is my preference because you can dose it precisely and the water stays clean.

The second option is the old method: drop a pinch of fish food in and let it rot. It works, because decomposing food releases ammonia, but you have no control over how much or how fast, the tank gets cloudy and smelly, and you often end up with fungus growing on the mush. I used this method on my first tank and it did the job, slowly and unpleasantly. If it is all you have, it is fine. If you can get bottled ammonia, get bottled ammonia.

Dose to roughly 2 to 4 ppm of ammonia, measured with your test kit. There is no benefit to going higher; very high ammonia can actually stall the process, and it grows a bigger bacterial colony than your eventual stocking will ever feed, which then partially dies back anyway.

Then you wait, and you test. For the first week or two, often nothing appears to happen. Ammonia sits stubbornly where you left it. Eventually it starts to fall as the first bacteria establish, and nitrite appears. When ammonia drops below about 1 ppm, dose it back up to your target. You are feeding the colony, and you keep feeding it throughout the process. A cycle that runs out of ammonia is a colony starving.

The timeline, and why rushing backfires

An unseeded fishless cycle usually takes four to six weeks. Sometimes it is done in three. Sometimes it drags on for eight, particularly in cool water or where the pH has drifted low. The nitrite stage is nearly always the long one; nitrite will spike to alarming levels and sit there for what feels like an eternity before the second group of bacteria catches up. This is normal. It is the point where most people panic, do something drastic, and set themselves back.

I understand the temptation to hurry. You have an empty tank in the living room and a mental list of fish. But every shortcut that skips the biology, rather than assisting it, just moves the ammonia spike to a point when fish are present. The bacteria multiply at their own pace. Products and tricks that promise a cycled tank in days are, at best, doing one of the legitimate seeding methods described below, and at worst doing nothing except lightening your wallet.

Testing: buy a liquid kit and trust it

You cannot cycle a tank blind, and you cannot do it properly with paper dip strips. The strips are imprecise at the low concentrations that matter, many brands do not test ammonia at all, and they degrade in the tub once opened. A liquid reagent test kit covering ammonia, nitrite and nitrate costs more up front and lasts for hundreds of tests. It is the one piece of kit I would never skip.

Test every day or two and write the numbers down. The pattern tells you exactly where you are:

  • Ammonia high, nitrite zero: the first bacteria have not established yet. Keep waiting.
  • Ammonia falling, nitrite rising: stage one is working. This is progress, even though the water is now more toxic than when you started.
  • Nitrite falling, nitrate climbing: stage two is establishing. You are on the home straight.
  • Ammonia and nitrite both zero, nitrate present: nearly there.

The finish line has a specific definition. Dose ammonia to 2 ppm, and if the tank reads zero ammonia and zero nitrite within 24 hours, with nitrate clearly present, the cycle is complete. Do a large water change to bring the accumulated nitrate down, match the temperature, and add your first fish within a day or two. Do not leave a cycled tank sitting empty and unfed for a week; the colony shrinks without food.

How to speed it up legitimately

There is one method that genuinely transforms the timeline: seeding with mature filter media. Squeeze a dirty sponge from an established filter into your new tank, or better, move a handful of ceramic media or a whole sponge across and put it directly in the new filter. That gunk is a live bacterial culture, and it can cut a six week cycle down to one or two, sometimes less. Between my own tanks I do this constantly; a new tub or quarantine tank gets a seeded sponge and is often processing ammonia within days. If you do not have your own established tank, a trusted friend’s filter or a decent local shop can sometimes oblige. Only take media from a healthy tank, since disease can hitch a ride with it.

Bottled bacteria products are the more heavily marketed option, and my honest experience is mixed. Some brands do seem to shorten the cycle by a week or so; others do nothing measurable. The bacteria in the bottle have been sitting on a warehouse shelf, and their condition on arrival is anyone’s guess. I treat them as a possible head start, never a substitute for testing, and never a licence to add fish early. If someone tells you a bottle makes a tank safe on day one, keep your money.

Fish-in cycling, only as a last resort

Sometimes you inherit fish with an uncycled tank, or the fish arrived before the knowledge did. Fish-in cycling can be done humanely, but it is a rescue operation, not a plan.

The fish themselves supply the ammonia, so your entire job becomes keeping concentrations low while the bacteria establish. That means testing daily, and doing a water change (30 to 50 percent, dechlorinated, temperature matched) whenever ammonia or nitrite creeps above about 0.25 ppm. In practice that is most days for several weeks. Feed lightly, stock as thinly as possible, and seed with mature media if you can get any.

The frustrating irony is that all those water changes remove the bacteria’s food along with the toxins, so a fish-in cycle is usually slower than a fishless one, on top of being stressful for both the fish and you. It works. I have done it. I would not choose it.

The mistakes that kill a cycle

A few errors come up again and again, and most of them involve accidentally killing the bacteria you spent weeks growing.

Rinsing filter media under the tap. Tap water contains chlorine or chloramine, added precisely because it kills bacteria. Rinse a mature sponge under the tap and you can wipe out a large slice of your colony in thirty seconds. Always rinse media in old tank water, in a bucket, during a water change. A gentle squeeze to clear the worst of the debris is all it needs; that brown coating is the colony itself, so do not scrub it back to pristine.

Chasing pH with chemicals. New keepers test pH, decide the number looks wrong, and start dosing adjusters. The result is a pH that swings up and down, which stresses fish and can stall the bacteria, who prefer stability and slightly alkaline conditions. Unless your water is genuinely extreme, leave the pH alone and choose fish that suit it. Hard, alkaline tap water is actually helpful during cycling because the carbonate buffering holds the pH steady, and plenty of superb species positively prefer it; I have written about hard water and the fish that thrive in it separately.

Replacing filter cartridges on schedule. Some manufacturers tell you to bin the cartridge monthly, which conveniently sells cartridges while throwing your bacteria in the bin. Sponges and ceramic media last for years. Rinse, reuse, and only replace media gradually, a portion at a time, when it is genuinely falling apart.

Impatience. Adding fish at the nitrite spike, doubling the ammonia dose to “hurry things along”, tearing the tank down because week three looks the same as week two. Almost every stalled cycle I have been asked about was not stalled at all. It was mid-process, and the owner intervened.

What a healthy cycled tank looks like

Once the tank is cycled and stocked, the drama disappears, which is exactly the point. Ammonia and nitrite read zero every time you test, permanently. The only number that moves is nitrate, which creeps up as the end product of the cycle, and managing it becomes the quiet rhythm of the hobby: a weekly water change of around 25 to 50 percent keeps it comfortably low, and live plants help by consuming nitrogen as they grow. Heavily planted tanks of mine barely register nitrate at all.

Add new fish gradually rather than all at once, since the colony needs a little time to grow into each increase in load, and keep the test kit handy for the first week after any change. Beyond that, the bacteria simply get on with it, unnoticed and unthanked, for as long as you keep their filter running and out from under the tap. Six weeks of patience at the start buys you years of tanks that just work.