A week-by-week calendar for your first freshwater month

A week-by-week look at what happens in the water from setup day to first stable readings, and how to dodge the standard new-tank mistakes that kill fish.

Cycling a new freshwater aquarium takes 3 to 6 weeks in most cases. A fishless cycle with dosed ammonia typically finishes in 4–6 weeks, a densely planted tank can carry light stocking after 2–3 weeks, and a fish-in cycle means 4–8 weeks of daily damage control. Seeding the filter with mature media from an established tank shaves a week or two off any of those routes.

Most of the harm done to a new tank happens in its first month, and almost all of it is well meant. The tank gets stocked early because the fish are exciting, water-changed often because clean water sounds like good care, fed generously because hungry fish look sad. Every one of those moves feels like attention. Every one can stall the process the tank needs to finish.

What follows is a calendar rather than a to-do list, because the same number means different things on different days. A small ammonia reading on day 14 is the cycle working. The same reading on day 30, with fish in the tank, is a problem. A 50% water change helps on day 24 and sets you back a week on day 9. Where you are in the month decides what the right move is.

Three honest ways to start

All three get you to the same destination: a filter that clears ammonia faster than your fish can make it. Where they differ is how long they take and how much risk the fish carry while you wait.

Fishless cycling with ammonia. Run the tank empty, dose pure ammonia to roughly 2 mg/L, and feed the bacteria for four to six weeks. This is the method to choose if you can. No animal is stressed while the colony builds, the cycle tends to finish deeper than a fish-in cycle does, and you come out with a clean record of what happened. The calendar below assumes this path.

Plant-cycled, sometimes called silent cycling. Plant the tank heavily from day one (at least 60% of the footprint in fast-growing stems) and add only a few animals. Live plants eat ammonia directly as fertiliser, and with enough leaf in the water, ammonia never builds up to a level your test kit can see. It's quicker, maybe two to three weeks before light stocking. The catch is that you can't verify it the way you verify a fishless cycle, and it punishes you quietly if the plant mass turns out too thin.

Fish-in cycling. Fish go in on day one and you chase the ammonia they produce with daily water changes. It does eventually work. The cost is that the fish spend weeks breathing low-grade ammonia and nitrite, which damages gills and shortens lives even when nobody dies outright. There's no version of this worth recommending. If you've already started here, skip to the rescue section at the end. The plant-cycled route compresses the calendar below into roughly its first fortnight with less testing; the principles don't change.

Day 1: setting up

Fill with dechlorinated water at the temperature the tank will run at. Substrate goes in first, then hardscape, then plants. The filter and heater run from the moment there's water to move.

A few day-one decisions are worth getting right, because they're awkward or impossible to undo later.

  • Match the substrate to the plan. A planted tank wants a nutrient-rich soil such as ADA Amazonia, Tropica Aquarium Soil, or JBL Manado. A fish-only or shrimp tank is happy on inert sand or gravel. Substrate is the one component you can't swap without tearing the tank apart, so decide before you pour water.
  • Test the source water on its own, first. Before you add a drop of anything, measure GH, KH, and pH on your tap or RO mix, plus copper if shrimp are anywhere in your future. Your starting water decides what will thrive: soft, acidic tap points you one way, hard, alkaline tap somewhere completely different. If the GH and KH labels confuse you, the difference between GH and KH sorts it out in a page.
  • Inoculate with mature media. A squeeze of dirty sponge from an established filter, a handful of seasoned substrate, or a wad of old floss in a media bag can shave a week or two off the whole thing. Bottled bacteria (Tetra SafeStart, Seachem Stability, Dr. Tim's One & Only) work, but inconsistently. Real seed material from a tank you know is healthy is the better bet by a wide margin.
  • Dose ammonia to 2 mg/L. Use plain liquid ammonia with no surfactants or perfume; read the label and shake it, foam means additives. Calculate the dose to bring the water column to 2 mg/L of NH₄. That's the bacteria's food for the next month and a half. Test the next morning to confirm you landed near 2.

No fish. No shrimp, no snails, nothing. The tank sits empty-looking for the next four weeks, and that's the point. The urge to "just test it with a couple of danios" is the same mistake as stocking the angelfish, only with cheaper victims.

Days 2 to 10: bacterial ignition

The first organisms to show up are usually Nitrosomonas, which turn ammonium into nitrite. Given warmth, oxygen, and surfaces to colonise they roughly double every twelve to twenty-four hours, but they start from almost nothing, so the visible result takes about a week to arrive. The shape of the readings:

  • Days 2 to 4: ammonia holds near 2 mg/L, nitrite at zero. Nothing's wrong. The colony is finding its feet.
  • Days 5 to 8: ammonia begins to slide, nitrite ticks up off the floor. First conversion is underway.
  • Days 8 to 10: ammonia heads toward zero and nitrite climbs hard, often straight past the top of your test scale. Whenever NH₄ drops below 0.5 mg/L, dose ammonia back up to 2. The bacteria need a steady supply to keep multiplying.

Leave the tank alone this week, in three specific ways. Don't do a water change; the colony seeding the water column matters too, and a change on day 7 walks you back to roughly day 4. Don't clean the filter; the brown film on the sponge is the biofilm where your bacteria live. And don't add fish, however tempting it gets around day 8 when ammonia reads zero. The second-stage bacteria aren't ready, and the nitrite they can't yet process is more dangerous to fish than the ammonia your filter just learned to handle.

Days 10 to 20: the nitrite peak

The second crew, Nitrospira in most freshwater tanks (older books credit Nitrobacter, but that's largely wrong for our conditions), converts nitrite into nitrate. They grow slower and start from an even smaller seed population. That lag is why nitrite is the headline number this week, frequently sitting well past 5 mg/L.

An off-the-scale purple nitrite reading is normal right now, not an emergency. There are no fish in the tank for it to hurt. Keep feeding the colony: dose ammonia back to 2 mg/L each time NH₄ falls under 0.5. By now the Nitrosomonas are mature enough to clear a fresh shot inside a day.

By the close of week two the picture usually reads like this:

Parameter What you see by ~day 20
NH₄ back to zero within 24h of each dose
NO₂ still high, but starting to fall
NO₃ first real readings, climbing slowly

This is the nitrogen cycle the textbooks name, happening in front of you over a fortnight. If you want the biochemistry underneath it, the nitrogen cycle in plain language lays it out without the jargon.

Days 20 to 28: settling

Around day 20 the second-stage bacteria have caught up. Nitrite drops toward zero. Ammonia stays at zero through each dose. Nitrate has been piling up for two weeks and now sits anywhere from 20 to 80 mg/L depending on how often you've fed the tank.

There's one test that tells you the cycle is finished. Run it like this:

  1. Dose ammonia to 2 mg/L.
  2. Wait a full 24 hours.
  3. Test both NH₄ and NO₂.
  4. Both should read zero.

Zero on both, a full day after a 2 mg/L shot, means the filter can already process the bioload of a modest stocking inside a day. You're ready for fish.

If ammonia comes back zero but nitrite is still showing, the second-stage colony is close but not finished. Give it another three to five days and run the test again. No need to intervene; just wait. If you want to sanity-check how far along you should be by a given day, the nitrogen cycle timeline maps the same milestones against the calendar.

Days 28 to 30: first fish

Do a water change before you stock anything. Make it a 50% change to pull nitrate back down to a sane starting point. The water-change impact calculator tells you exactly where a given dilution lands so you're not guessing.

Then stock no more than a quarter of the final plan. A 100-litre tank destined for twenty cardinal tetras gets five fish on day 30, not twenty. The cycle is done, so that's not the worry. The issue is colony size. Your bacteria grew to match the ammonia you were dosing, not the load twenty fish will produce. Drop the whole inventory in at once and you can push ammonia past what the current colony clears, which gives you a mini-cycle: a brief ammonia spike while the bacteria scale up.

Wait about two weeks, then add the next quarter, then the next. Somewhere around six to eight weeks after the first fish you're at full stocking, and the colony has grown alongside the load the whole way. Slow suits this part of the hobby, which is roughly Manfred's view of most of it.

Mistakes that reset the clock

The habits that hurt a cycling tank hardest aren't the careless ones. They're the maintenance routines people learn from established-tank advice and apply far too early.

  • Rinsing filter media in tap water. Chlorine and chloramine kill the bacteria you spent a month growing. On a mature tank you rinse media in old tank water during a change. During cycling you don't rinse it at all.
  • Antibiotics and antibacterials. Anything ending in "-mycin", products like Furan-2, methylene blue at treatment strength: all of them gut the filter. Sick new arrivals get treated in a hospital tank, never the main one.
  • Deep substrate cleans. Driving a gravel vac deep during cycling tears open an oxygen-poor pocket of rotting organics and ammonia, which can briefly shove NH₃ over the toxic line. A light surface vacuum is fine. Save the deep digs for month two.
  • Overfeeding. Five cardinals do not need a hearty pinch twice a day; five small fish eat about what fits on a fingernail. Anything more rots, ammonia jumps, and the colony scrambles to keep pace. The feeding-to-nitrate calculator shows how much nitrate a given gram of food becomes, and the same arithmetic runs upstream on the ammonia side.

What if there are already fish in here?

If you're reading this on day 9 of a tank you stocked on day 1, breathe. It's recoverable. Don't rehome the fish in a panic, and don't tear the tank down.

  1. Test ammonia and nitrite every day. Anything over 0.25 mg/L of either counts as acute stress and needs action.
  2. Water-change to dilute. A 50% change roughly halves the worst reading. Do it again the next day if the numbers are still high.
  3. Cut feeding right back. Once every other day is plenty. Less food in means less ammonia to deal with.
  4. Use an ammonia-binding conditioner. Seachem Prime locks ammonia and nitrite for around 48 hours into a form the bacteria can still eat but the fish can't be poisoned by. Redose every two days until the tests settle.
  5. Then wait. The cycle still finishes, just on the schedule the fish set rather than the one you'd have chosen. Expect four to eight weeks of attentive management instead of a clean fishless run.

The fish you already have will most likely pull through. The ones you were planning to add in week two should stay out until the tests read clean for seven days straight. Log those daily numbers somewhere the trend is visible at a glance; scattered sticky notes won't tell you whether the curve has actually turned or you're just hoping it has.

Manfred

Manfred quietly remembers every test, dose, and water change you log. The trends fall out — no spreadsheet required.

Open a free reefnotes account →