Has my aquarium cycle crashed? Signs, causes, what to test and what to do

What fishkeepers mean by a cycle crash

"Cycle crash" (sometimes "mini-cycle" or "cycle disruption") is hobby shorthand, not a term with one fixed scientific definition or a lab test that confirms it. It's generally used to describe a meaningful loss of the tank's ammonia- and nitrite-oxidising bacterial population — the same organisms responsible for the nitrogen cycle — severe enough that ammonia and/or nitrite start accumulating in a tank that had previously tested at zero.

The scale varies enormously between cases people describe this way. Sometimes it's a small, short-lived dip that clears within days; sometimes it's a much larger loss that takes weeks to recover from, closer to restarting part of the original cycling process. Both get called a "crash" in hobby conversation, which is part of why the term is more useful as a prompt to investigate than as a precise description of what happened.

What the test results may look like

A genuine disruption typically shows as a detectable ammonia and/or nitrite reading in a stocked tank that has previously tested at zero, sometimes alongside a nitrate reading that's lower than usual because less ammonia is being converted all the way through. But a positive reading by itself doesn't distinguish between a genuinely reduced bacterial population, a sudden spike in waste that has temporarily outpaced a perfectly healthy population, and a mistaken or expired test kit. All three can produce a very similar-looking number.

Confirmed detection, inconclusive results, normal variation, or testing error?

Before reacting to a single result, it helps to place it in one of these four categories. None of them can be identified from one test alone — retesting and a little context is what actually tells them apart.

Common causes

More than one of these is often involved at once, and they fall into a handful of natural groups rather than being one long list of unrelated possibilities.

The most common group is simply too much waste for the bacteria to keep up with: new livestock added faster than the existing population can adjust to, a tank that's quietly grown to hold more than it comfortably supports, overfeeding where uneaten food decays and adds directly to the ammonia load, or a dead and decomposing fish, snail or shrimp that's gone unnoticed. None of these involve the bacteria being damaged at all — there's just suddenly more for them to process than there was before.

A second group is mechanical: reduced filter flow from a fouled impeller, a partially blocked intake, or a filter that wasn't primed properly after maintenance, and extended power loss, which stops the filter and starves the bacteria of the oxygen and water movement they need. Give either of these long enough and the population itself starts to suffer, even though nothing was done to it directly. If you're in the middle of an outage right now, see our guide to protecting an aquarium during a power outage for what to do before it gets to this point.

A third group is more direct damage to the bacteria themselves, usually from well-intentioned maintenance[1]. Rinsing filter media in untreated tap water exposes it to chlorine or chloramine, which is added specifically to kill microorganisms. Scrubbing or replacing media far more thoroughly than needed, or swapping out sponges and cartridges that were carrying most of the bacterial population, has much the same effect. Medication and other chemicals can do it too — copper-based treatments in particular are known to inhibit nitrifying bacteria at low concentrations[4], and some other medications carry similar warnings on their own packaging.

Major substrate disturbance or a wholesale substrate change belongs in this same group and is worth naming explicitly, since it's a common, planned event that people don't always connect to cycle stability beforehand. Substrate carries a genuine share of a tank's nitrifying bacteria, particularly in an established tank with deep or long-undisturbed substrate, and digging it all out, replacing it entirely, or deep-cleaning it all at once removes a meaningful part of that surface area in one go — on top of whatever gets stirred into suspension and temporarily fouls the water. See planning a substrate change without crashing the cycle below if this is something you're about to do.

And sometimes what's being called a "crash" isn't really bacterial loss at all. A testing error — an expired kit, a misread colour chart, contaminated glassware — can produce a reading that looks identical to a genuine one. A recently established aquarium that was never fully cycled in the first place can also get mistaken for a crash, when it's really an original cycle that hadn't finished establishing. And not every dip is total: a temporary imbalance, a partial and short-lived reduction in capacity, is a very different situation from the population being wiped out, even though both get described the same way in hobby conversation.

Cloudy water, algae growth, a zero nitrate reading or unusual livestock behaviour can all prompt someone to suspect a crash, but none of these confirms one on its own — each has other common explanations, and the only way to know is to test ammonia and nitrite directly.

Why does my aquarium cycle keep crashing?

A single crash usually has one identifiable trigger. A tank that crashes repeatedly is a different situation worth treating separately: it almost always means either the same mistake is being repeated, or an underlying condition was never actually resolved the first time — not that the bacteria are somehow fragile or that the tank is simply unlucky. Working through the causes above with an eye for what repeats between incidents, rather than treating each one as an unrelated event, is usually what actually finds the fix.

• Repeated filter-media replacement or excessive cleaning. If a detectable reading keeps reappearing shortly after routine maintenance, the maintenance itself is the likely repeating trigger — replacing more media than necessary each time, or cleaning so thoroughly that the population never fully re-establishes before the next round.
• Biological media rinsed in untreated tap water. A one-off rinse under the tap can meaningfully reduce the population; doing it as a matter of habit every time media looks dirty produces the same disruption on a repeating schedule. Rinse only in removed tank water.
• Chronic overstocking or overfeeding. A single overfeed produces a brief, one-off spike. An ongoing feeding routine or a stocking level that consistently sits close to what the tank's bacteria can process leaves little margin — the population never gets meaningfully ahead, so it doesn't take much (a slightly larger feed, a slightly warmer week) to tip it into a detectable reading again and again.
• Recurring filter-flow, pump or power problems. A single short outage is unlikely to cause lasting harm, but an intermittent pump fault, a filter that keeps losing prime, or repeated brief power interruptions can each starve the bacteria of oxygen and flow on a repeating basis, producing a pattern of readings rather than one isolated event.
• Medication or chemical effects, where properly evidenced. Copper-based treatments are specifically documented to inhibit nitrifying bacteria at low concentrations[4]; a treatment course repeated or extended without accounting for this can produce a similar disruption each time it's used. We're not aware of comparably specific evidence for most other medications, so treat this as a documented risk for copper specifically rather than an assumption about any product.
• A tank that never actually completed its first cycle. What looks like repeated crashing can sometimes be a tank that was stocked before the bacterial population was genuinely established in the first place — each "crash" is really the same unfinished cycle struggling under normal bioload, not a new disruption each time. See our decision guide to whether a tank is genuinely cycled if you're not confident the original cycle ever properly finished.
• Testing errors repeatedly mistaken for crashes. An expired kit, a degraded reagent, or a consistently misread colour chart doesn't fix itself between tests — it can produce the same false-positive reading over and over. If "crashes" keep happening on the same kit with no other explanation, testing the kit itself (a fresh reagent set, a second kit, or a known-good sample) is worth doing before assuming the bacteria are at fault again.
• Recovery actions causing a second disruption. This one is easy to miss: reacting to a first crash with an unusually thorough clean, a full media swap, or sterilising surfaces "to be safe" can itself cause the second crash, especially if it happens again after the next reading. See "what not to do" below — the safer response to a confirmed reading is targeted (reduce feeding, measured water changes, leave the filter alone), not a bigger version of the maintenance that may have caused the first one.

Tank history is usually more revealing than any single reading here: noting what maintenance, stocking or feeding changes preceded each incident, and comparing them once there's more than one data point, tends to expose a repeated trigger that a single test result never could. It's also worth being direct about what a reading does and doesn't prove: a detectable ammonia or nitrite result confirms that waste is currently outpacing what's being processed at that moment — it does not, by itself, confirm that the bacterial population has collapsed. As the table above sets out, the same reading can also come from a genuine but partial dip, a sudden waste increase acting on an otherwise healthy population, or a testing error, and treating every occurrence as proof of a full bacterial die-off risks a bigger reaction (and more disruption) than the situation actually calls for.

Planning a substrate change without crashing the cycle

Swapping substrate — changing sand for gravel, adding aqua soil, or simply replacing old, worn-out material — is a normal thing to want to do, and it doesn't have to mean restarting your cycle from scratch. It does need planning, because the substrate itself is carrying part of your biological filtration, not just the filter. The following reduces the risk; it doesn't eliminate it, and even a carefully staged change can produce a temporary reading worth testing for afterwards.

• Don't combine it with other disruptive maintenance. Replacing substrate and deep-cleaning or replacing filter media in the same session removes bacteria from two surfaces at once, right when the tank needs as much of its existing population intact as possible.
• Leave the filter alone and keep it running and oxygenated. Your filter media is likely carrying more of the bacterial population than the substrate is — protecting it matters more here than during almost any other kind of tank work.
• Don't let removed biological media dry out. If you do need to lift or temporarily remove filter media or substrate you intend to reuse, keep it wet in tank water throughout — drying kills the bacteria living on it.
• Work in stages where practical. Replacing a portion of the substrate at a time, over several sessions, gives bacteria on the remaining, undisturbed substrate and in the filter a chance to cover the gap. This isn't always possible — a full substrate-type change (sand to gravel, for example) or a hardscape rebuild often can't sensibly be staged, and a complete change done carefully with the rest of this list followed is still far lower-risk than combining it with other disruption.
• Match and condition the replacement water. New substrate needs rinsing and the tank needs refilling with dechlorinated, temperature-matched water, the same as any other significant water change — see our water changes guide.
• Monitor ammonia and nitrite afterward. Test daily for at least the following week or two, the same way you would after any other suspected disruption.
• Reduce feeding and delay new livestock until readings are stable. Less waste for a temporarily reduced bacterial population to process, and no additional bioload until you've confirmed it's coping.

It's worth being direct about one common assumption: old tank water alone does not preserve your biological filtration. As explained in our decision guide to whether a tank is genuinely cycled, the bacteria that matter live on surfaces — filter media, substrate, décor — not free-floating in the water column, so keeping the same water while removing the surfaces it used to flow past doesn't protect the population living on them.

What to test and retest

• Ammonia and nitrite together — these are the two readings that actually indicate a biological-filtration problem. See our ammonia and nitrite guides for what a specific reading means.
• Nitrate — a sudden drop can be a secondary clue that less ammonia is making it all the way through the cycle, though it's not diagnostic by itself.
• pH — nitrifying bacteria are pH-sensitive, and a significant, sustained drop can itself slow or stall biological filtration[5]. See our pH guide if your pH has fallen noticeably or keeps falling.
• Retest daily while a reading is detectable, using a fresh sample each time, rather than relying on a single test to decide what's happening.
• Cross-check an unexpected result with a second sample, a fresh kit, or (if available) a different test method before treating a borderline reading as confirmed.

What to do when livestock are present

Reduce immediate exposure

1. 1

   Do a measured water change

   Dilutes whatever ammonia or nitrite is currently in the water. Use appropriately conditioned, temperature-matched water. The Water Change Assistant can work out the volume for your tank and show the likely effect on your reading.

2. 2

   Pause or reduce feeding

   Less food in means less ammonia produced while the population recovers.

3. 3

   Increase surface agitation or aeration

   Supports both livestock and the recovering bacteria, which also need oxygen.

Investigate and hold steady

1. 4

   Look for a specific recent cause

   Recent filter maintenance, a power cut, new medication, new livestock or a possible dead animal are the places to check first.

2. 5

   Hold off adding further livestock

   Wait until ammonia and nitrite are both back to zero and staying there.

3. 6

   Retest daily until stable

   Keep going until both readings hold at zero, rather than stopping after the first improvement.

What to do during fishless cycling

If no livestock are present, there's no welfare urgency — a disrupted or restarted cycle is inconvenient, not dangerous to animals. Identify what disrupted it (see the causes above), then continue dosing ammonia and testing according to whichever established cycling method you're following. There's no shortcut that reliably speeds up bacterial regrowth beyond what your chosen method already describes; consistency in testing and dosing matters more than trying something new mid-process. See our nitrogen cycle guide for how to track progress and recognise when a tank is genuinely cycled, rather than assuming it from a calendar date.

What a water change can and cannot do

A water change genuinely helps here, but it's worth being precise about what it's actually doing, since it's easy to credit it with more than it delivers. It dilutes the ammonia or nitrite already in the water, directly and immediately — that part is straightforward. What it doesn't do is reduce how much new waste is entering the system (that's a feeding and stocking question), preserve what's left of your biological filtration (that comes from leaving the filter alone), or investigate the underlying cause. Most importantly, it doesn't rebuild the bacterial population itself: that happens through its own biological process over days to weeks, given stable conditions and an intact surface for the population to regrow on, not through dilution[2]. Treat the water change as the safety measure that buys time for that recovery, not as the recovery itself.

If you calculate a second, identical water change using the Water Change Assistant, treat that figure as a mathematical comparison of what repeating the same change would do on paper — not as an instruction to perform a second change regardless of your next actual reading.

Supporting recovery without destroying more bacteria

• Leave the filter running continuously, even if flow seems reduced — switching it off removes oxygen and flow the recovering bacteria need.
• Avoid cleaning or replacing any more filter media while readings are elevated.
• Rinse anything you do need to clean in removed tank water, never tap water.
• Keep feeding light and consistent rather than skipping days entirely and then overfeeding to compensate.
• Some products marketed as concentrated nitrifying-bacteria supplements may help seed a recovering system, but their effect varies by product and by what's already re-establishing naturally — testing ammonia and nitrite directly is still the only way to confirm recovery, regardless of what you've added.

What not to do

A few common reactions can do more harm than the original problem, and most of them share a theme: doing too much, too thoroughly, all at once. Replacing all the filter media at once is one of the most common ways a recovering, or even an already-established, population gets set back a second time — and sterilising the filter, or deep-cleaning every surface in the aquarium at the same time, has much the same effect, since substrate and décor also carry part of the bacterial population. However tempting a really thorough clean feels right now, the safer instinct is to leave as much of that surface area undisturbed as you can.

A few other habits are worth resisting too. New livestock shouldn't go in while ammonia or nitrite is still detectable, even if your current animals seem to be coping. Stacking multiple "quick fix" additives without understanding what each one actually does tends to complicate the picture rather than clarify it, and a bottled-bacteria product's own claims aren't a substitute for testing — confirm recovery with your kit, not with the packaging.

Chasing a pH number is another common misstep: pH that keeps drifting down despite water changes usually points to low buffering capacity (alkalinity, or KH) rather than something a pH adjuster alone will fix — see our pH guide for more on this. And if you've used the Water Change Assistant to work out what a second, identical water change would do, treat that as a mathematical comparison rather than a fixed instruction to follow regardless of what your next actual reading shows.

When livestock distress needs urgent specialist or veterinary help

Watch for gasping at the surface, rapid or laboured gill movement, clamped fins, sustained loss of appetite, or unusual lethargy — these can indicate ammonia or nitrite exposure[3], but also other water-quality or health problems, so treat them as a prompt to test rather than a diagnosis in themselves. If losses have already occurred and continue, see our guide on why fish sometimes die one by one.

Frequently asked questions

Sources and editorial information

Good to know: This guide provides general informational guidance only. It is not a veterinary service and does not replace specialist or aquatic health advice for serious livestock health issues.

• [1] University of Florida IFAS Extension — "Ammonia in Aquatic Systems" (FA16/FA031) Used for the nitrification process, where nitrifying bacteria colonise (including substrate), and how filter-media handling affects the population.
• [2] Hovanec & DeLong (1996), "Comparative Analysis of Nitrifying Bacteria Associated with Freshwater and Marine Aquaria," Applied and Environmental Microbiology Cited for background on which organisms actually establish in aquarium biofilters, relevant to why regrowth after a disruption isn't instant.
• [3] Merck Veterinary Manual — "Environmental Diseases of Aquatic Animals in Aquatic Systems" Used for clinical signs of ammonia and nitrite exposure and general water-quality management.
• [4] University of Florida IFAS Extension — "Use of Copper in Marine Aquaculture and Aquarium Systems" (FA165) Used for the specific concentration at which copper treatments inhibit ammonia and nitrite oxidation in a biofilter.
• [5] California Water Environment Association — "How Alkalinity Affects Nitrification" (Evans & Sober) Used for the finding that nitrification rates decline significantly at low pH, cited here for the general water-treatment chemistry rather than an aquarium-specific figure.

Related guidance