05 Food
Fasting, rest days and when not to feed
Several pages on this site tell you to fast the fish for a few days. This is the page explaining why keepers do it, how long is safe, and the several cases where withholding food is the wrong instruction entirely.
Several pages on this site tell you to stop feeding for a few days: the buoyancy page, the constipation page, the overfeeding page, the transport section of the handling guide, and the emergency response to an ammonia reading. None of them explains why keepers do it, how long is safe, or the cases where it is precisely the wrong thing to do.
What a fast actually does
A goldfish has no stomach. As the nutrition page sets out, food passes along the gut rather than waiting in a holding chamber.
Three effects are usually given as the reasons a fast helps. The first two are keepers’ explanations; the third is straightforward:
The tract clears. With nothing going in, the gut empties over a day or two, and anything that was moving slowly moves on.
Pressure may come off the swim bladder. The gut runs close against the swim bladder, and in the deep-bodied fancy breeds the whole abdomen is compressed into a shorter space, so a full tract may press on the organ that controls buoyancy. That is the usual explanation and it has not been tested; keepers report that some fish right themselves once the gut empties, and infection is among the other causes of buoyancy trouble.
The tank’s nitrogen load drops. Almost all the ammonia in an aquarium originates in food — through the fish’s waste and through anything uneaten. Stop feeding and the input stops with it, which is why a fast is part of the emergency response to a water quality problem rather than only a treatment for the fish.
Why goldfish tolerate it so well
Because they are temperate animals whose wild ancestors ate little or nothing in winter, living partly on reserves laid down in autumn. The seasonal feeding page covers the annual version of that, and a pond goldfish eats very little through every winter.
So a healthy, well-conditioned adult can generally go a few days with nothing, and up to about a week in a cool room, though it loses condition the longer it goes. This is a genuine biological margin, but a limited one.
What it does not license is vagueness. The fish that copes easily is a well-conditioned adult in stable water. A thin fish, a juvenile, or a fish already fighting something has much less to draw on.
The situations where fasting is the answer
A buoyancy problem. The usual first step for a goldfish floating, rolling or sinking is two to three days with no food. Keepers report that it helps some fish, though there is no controlled evidence and infection is among the causes, and it costs a healthy adult little; if three days of fasting changes nothing at all, the cause may not be the gut.
Constipation and long trailing faeces. Same approach, same duration. The constipation page covers what the faeces are telling you.
Overfeeding, caught in time. A tank where the fish are visibly overfed and the water is suffering benefits from a few days off while the maintenance catches up. The overfeeding page is the fuller argument.
An ammonia or nitrite reading above zero. Cut feeding right back, or stop for a few days, while water changes bring both readings down. Water changes remove what is there; reducing food slows it being replaced. Do not fast for weeks waiting for zero: in a tank that has not cycled the readings can stay above zero for that long, and the water changes are the real answer.
Before a journey. Twenty-four hours without food before transport means far less fouling of the bag or container, which matters more than the meal does. The handling and moving guide covers the rest of the preparation.
Before and during treatment. A fish under medication is often not eating properly anyway, and uneaten food in a hospital tank you cannot clean normally is the fastest route to an ammonia problem on top of everything else.
During a heatwave, or an oxygen problem. Digestion itself consumes oxygen. A fasted fish has more margin, and a tank where the fish are gasping is not a tank that needs another meal in it. The aeration guide covers the fix; reducing food is the free part of it.
What a fasting fish does
Normal, and worth expecting so that it does not look like distress.
It forages harder. Sifting the substrate, picking at algae, working over the decor and the plants. A goldfish forages constantly at the best of times and does it more when nothing arrives on schedule. This is ordinary behaviour, not desperation.
It begs more. Rushing the front of the tank whenever you appear. Goldfish learn feeding routines quickly and notice immediately when one stops.
It eats your plants. The practical cost of a longer fast in a planted tank. Soft-leaved species take the brunt of it, and a fish fasted for a week will do real damage to them.
It slows down, a little, after several days. Less activity and less interest is normal in the second half of a long fast and should reverse within a day of feeding resuming. What is not normal is a fish that stays listless once it is eating again, which is a different problem that happened to coincide.
None of this is a reason to shorten a fast that has a purpose. All of it is a reason to tell anyone else in the house what you are doing, since a common way for a therapeutic fast to fail is that somebody takes pity and feeds the fish on day two.
Holidays, and the longest fast most fish will meet
The longest fast a typical goldfish meets is not therapeutic at all — it is the days or weeks while the household is away, and it is the version where the limits matter most.
A healthy, well-conditioned adult can generally go a few days with nothing, and up to about a week in a cool room, but an aquatic vet’s emergency limits shrink to three or four days at 21–24 °C, and she calls them guidelines for an emergency and not an excuse for holiday care. For anything longer than a few days, someone should feed the fish. The leaving them while away guide covers the whole arrangement; the feeding part reduces to three points:
Do not overfeed before you go. A large meal the morning of departure produces a fouled tank on day two and a fish sitting in it until you return. It is a common mistake on the way out of the door.
Do not use a dissolving block feeder. They dissolve unevenly, many goldfish ignore them, and what is left degrades in the water.
Be careful who you ask. A neighbour given an open tub and no instruction will feed generously, because that is what care looks like to someone who does not keep fish. If you do ask someone, leave pre-measured portions for every second day and say plainly that the fish will be fine if they skip one and will not be fine if they double one.
The filter fasts too
A subtle consequence of any long fast, and one that catches people after a holiday or an ammonia crisis.
The bacteria in a filter are sustained by the ammonia the fish produce, which comes from food. Take the food away for a long stretch and the colony can shrink to match the reduced supply. Return and feed a full ration immediately, and for a few days the filter may be processing a load larger than the population it currently supports.
So resume gradually after any fast longer than about a week: a small feed the first day, building over three or four days, with the water tested through that period. It is the same logic as restocking a tank slowly, and the same logic that the cycling guide applies at the other end.
How long
| Duration | Verdict |
|---|---|
| One day | Well within what a healthy adult tolerates |
| Two to three days | The standard therapeutic fast. A healthy, well-conditioned adult copes |
| Up to about a week | The most a well-conditioned adult is generally asked to manage, in a cool room; fewer days in warm water. Not for fry or a thin fish |
| A week or more | Only with a reason, and not as a holiday plan; condition is lost the longer it goes |
| Weeks | Only in cold water, below about 12 °C, where the fish eats little of its own accord |
Longer is rarely better. If three days of fasting has not changed a buoyancy problem, a fourth and fifth day are unlikely to, and the useful next step is to look at something other than the gut.
When fasting is the wrong answer
This section matters more than the last one, because withholding food from a fish that needs it is a real way to do harm while believing you are helping.
A fish that has already stopped eating. It is not in need of a fast; it is not eating, and the question is why. Loss of appetite has its own causes — water quality, temperature, an internal problem, or a fish being outcompeted — and none of them is improved by withholding food it was not taking anyway.
A thin fish, or one losing condition. A visible ridge along the back, a head that stands proud of the body, a fish that has narrowed from above. It has little to fast on, and what it needs is the cause found rather than a fast.
Fry and juveniles. A young goldfish has little reserve and needs frequent small feeds. A fry or juvenile held short of food may stay smaller than well-fed siblings; fish often catch up partly or fully when feeding improves, but not always, so do not fast them.
A fish with a suspected internal parasite. Medicated food is the usual route for treating internal parasites, and it needs a fish that is eating. Fasting removes that route, and a fish that has stopped eating on its own cannot take medicated food at all, though a vet can use baths or injections — which is why a declining appetite is worth acting on early.
A chronic, structural buoyancy problem. Some deep-bodied fancies seem to have a conformation that produces recurring buoyancy trouble regardless of feeding. A fast may give temporary relief, and repeating it every fortnight is not a management plan. Keepers report that sinking food, soaked and smaller, more frequent meals help some fish, though there is no controlled evidence, and realistic expectations matter.
A water quality problem you have not fixed. Fasting reduces the input while you deal with it. It is not the fix, and a tank that needs permanent fasting to keep its ammonia down is a tank that is overstocked or under-filtered.
The weekly rest day
Widely recommended, genuinely optional, and worth being clear about.
For a goldfish on a good diet in a well-maintained tank, one day a week with no food does very little. The fish is not accumulating anything that needs clearing, and the nitrogen saving is a seventh of not very much.
Where it earns its place is as a discipline for the keeper. Many people overfeed and feed by eye, and a fixed day when nothing goes in is a simple brake that does not rely on judgement.
Two cautions. It does not license heavier feeding on the other six days, which is the obvious way to cancel it out. And it is of no use at all in a tank where the underlying pattern is too much food — the feeding calculator and the how often to feed guide address that directly.
Breaking the fast
This is where people undo the whole thing, and it takes ten seconds to get right.
Small. A fraction of a normal portion. The gut has been empty for three days and a full meal takes it from empty to overfilled in one sitting.
Soaked, if you use dry food. Dry food expands as it absorbs water, and many keepers soak it first so that it does not swell inside the fish; there is no controlled evidence for this, but it costs nothing. Thirty seconds in a spoonful of tank water is enough.
Sinking, not floating. Many keepers prefer sinking food after a fast so that a hungry fish does not gulp air at the surface; again there is no controlled evidence, but it is a cheap precaution. The sinking versus floating page covers why keepers lean this way more for fancies than for single-tails.
Then build back over two or three days rather than returning to a full ration immediately.
If the problem returns within a week of normal feeding, the fast was not the answer — look at the feeding regime, and at the other causes of buoyancy trouble, infection included.
Fasting in a community tank
You cannot fast one fish among others. Goldfish forage continuously, and food that settles for the others will be found by the fish you were trying to keep off it. Feeding the rest while one watches is not a fast; it is a slower one.
If a single fish genuinely needs several days with no food, it needs to be somewhere else. This is one of the non-medical uses of a hospital tank — no treatment, no salt, just a bare tank, clean water and the ability to control what that fish eats.
The reverse case is worth noting too: a fish that appears to need fasting may simply be the one that eats fastest, and a tank where one fish takes most of the food has a feeding-distribution problem rather than a gut problem.
Cold water makes it moot
Below about 10 °C a goldfish eats little of its own accord and digestion has slowed steeply, so it is close to fasting whatever you do. A therapeutic fast in cold water adds little to a fish that is already eating next to nothing, and a buoyancy problem in a cold pond is worth treating as a possible illness rather than a gut to be cleared.
The corollary matters more: if you are treating a buoyancy or gut problem, the fish needs to be warm enough to eat and digest before and after the fast, which is one of the arguments for dealing with a sick pond fish indoors rather than waiting for spring.
In summary
| Situation | Fast? | How long | Then |
|---|---|---|---|
| Floating, sinking or rolling | Yes | 2–3 days | Small portion; soaked, sinking food if you use it |
| Long trailing faeces | Yes | 2–3 days | Reduce ration; add vegetables |
| Ammonia or nitrite above zero | Cut right back | A few days, while water changes bring the readings down | Resume lightly |
| Before transport | Yes | 24 hours | Feed on arrival, lightly |
| During treatment | Light or none | The course | Build back slowly |
| Heatwave or low oxygen | Reduce | Until it passes | No catching up afterwards |
| Fish already off its food | No | — | Test water, find the cause |
| Thin fish, or one losing weight | No | — | Find the cause |
| Fry or juveniles | No | — | They need frequent small feeds |
| Suspected internal parasite | No | — | Medicated food needs a fish that is eating |
The through-line: fasting is a cheap first step for a gut that may be overloaded and a bad tool for most other things. When it helps it helps within three days, and if it does not, the cause is probably somewhere other than the food.
Frequently asked questions
How long can a goldfish go without food?
A healthy, well-conditioned adult goldfish can generally go a few days without food, and up to about a week in a cool room, which is the figure a University of Florida fish vet gives. An aquatic vet who sets out emergency limits ties them to temperature: three to four days at 21–24 °C, four to five at 18–21 °C, five to six at 15–18 °C and six to eight at 12–15 °C, with one to three days above 24 °C and weeks below 12 °C, and she calls them guidelines for an emergency and not an excuse for holiday care. Goldfish are temperate fish whose wild ancestors ate little or nothing in winter and lived partly on reserves built in autumn, which helps explain the margin, but it is a limited one. Fry, juveniles, and any fish that is already thin have much less.
Why do keepers fast a goldfish that is floating or sinking?
Keepers report that two or three days without food helps some fish that are floating, rolling or sinking. The usual explanation is that a full or slow-moving gut presses on the swim bladder, which lies beside it, and that the pressure eases when the gut empties. There is no controlled evidence for that explanation, and infection is among the causes of buoyancy trouble, so a fish that does not improve needs a closer look rather than a longer fast. The fast costs a healthy adult little, which is why it is the usual first step, and what you feed afterwards matters more than the fast itself.
Should I give my goldfish a fasting day every week?
It is a reasonable habit and not a necessity. For a fish on a good diet in a well-maintained tank it adds little; for a keeper who tends to feed generously — which is many of us — a fixed day with nothing going in is a useful brake and slightly reduces the weekly waste load. Treat it as a discipline for the person rather than a health intervention for the fish, and do not let it license heavier feeding on the other six days.
My goldfish has stopped eating. Should I fast it?
No. A fish that has stopped eating is not in need of a fast — it is already not eating, and the question is why. Withholding food from an animal with no appetite adds nothing and delays finding the cause, which may be water quality, an internal problem, a temperature that has dropped, or a fish being outcompeted at feeding time. Test the water and the temperature the same day, then work through the loss-of-appetite page.
Does the pea trick actually work?
It is traditional and keepers report that it helps some fish, but there is no controlled evidence that peas relieve constipation or buoyancy problems, and the usual explanation, fibre and moisture in a gut that is moving slowly, has not been tested. A small deskinned blanched pea is a harmless thing to try. The caveat is that a pea is nearly always given at the end of a fast, so the fast may be doing the work — and if the problem recurs the answer is a change in how the fish is fed rather than a pea every fortnight.
Should I stop feeding if my ammonia reading is high?
Yes — cut feeding right back, or stop for a few days, while water changes bring the readings down. Food is where almost all the ammonia in a tank comes from, both through waste and through anything uneaten, so reducing it removes the input while water changes deal with what is already there. Do not fast for weeks waiting for zero: in a tank that has not cycled the readings can stay above zero for weeks, and the water changes are what bring them down. A few days without food costs a healthy goldfish little; a few days at a measurable ammonia level costs it a great deal.
How do I break a fast without starting the problem again?
Small first, and for a fancy many keepers also soak the food and choose a sinking one. A large meal straight after a fast takes the gut from empty to overfilled in one sitting, and keepers report that a buoyancy problem that had settled can come straight back. Soaking and sinking food are keepers' practice rather than something tested: the idea is that the fish swallows less air and nothing dry swells inside it, and there is no controlled evidence for either. Give a fraction of a normal portion, then build back over two or three days.
Can I fast one fish in a community tank?
Not really, which is a practical problem people run into. Feeding the others while one fish watches does not fast it — goldfish forage constantly and will find what settles. If one fish genuinely needs to be off food for several days, it needs to be somewhere else, which is one of the non-medical uses of a hospital tank.