01 Care guide
Using medications safely
Many goldfish lost during treatment are lost to the treatment and not to the disease: a dose calculated on a tank's advertised volume, a filter that was never opened, or two medications used in the same week.
Nearly every page in the health section ends with some version of “treat it”. This is the page about what that actually involves, because treating a goldfish is where a manageable problem most often becomes a dead fish.
The pattern is consistent. The dose is calculated on the number printed on the box the tank came in rather than the water actually in it. The carbon stays in the filter and quietly removes the medication. The aeration is not increased. A second product goes in three days later because the first one “did not seem to work”. None of that is carelessness — it is what happens when the instruction is “treat it” and nobody explains the rest.
Before anything: should you be medicating at all?
The great majority of goldfish health problems are environmental in origin. That is the whole argument of the diseases overview: many of the organisms behind these conditions are already present in most tanks, and they become a problem when the fish’s defences are down — usually because of ammonia, nitrite, a crash in temperature, or a tank too small for the animal in it.
Medication does nothing about any of that. Worse, most treatments are a stress in themselves, so adding one to a fish that is already struggling in bad water makes the fish’s position worse rather than better.
So the order of operations is fixed:
- Test the water. Ammonia, nitrite, nitrate, pH, temperature. The testing guide covers how, and this step is not optional — if ammonia is above zero you have found your answer and no medication addresses it.
- Correct what the test found. Water changes, and whatever the reading points at.
- Identify what you are treating. Not “it looks ill” but a specific sign, which the health section is organised around.
- Then, if a treatment is genuinely indicated, medicate properly.
A great many fish recover at step two.
Your real water volume
This is the most consequential number on the page and almost nobody has it right.
A tank sold as 100 litres is not holding 100 litres of water. It is filled to a few centimetres below the rim. Gravel displaces a surprising amount — a 5 cm bed across a 100 × 40 cm footprint fills 20 litres of the tank, and roughly 12 of them are stone rather than water. Rocks, wood, a large ornament, the filter body and the heater all add to it. Between the fill line and the furniture, the real figure is commonly 10 to 25 percent below the nominal one.
Dose on the nominal figure and you are overdosing by that margin, every time, on a fish that is already unwell.
Measure it. Internal length × width × filled depth in centimetres, divided by 1,000, gives litres. Then subtract a realistic allowance for substrate and decor. If you want the arithmetic done for you, the volume calculator does it:
Write the answer down and tape it to the tank. You will need it again — for this course, for the next water change, for salt, and at three in the morning when something has gone wrong. It is a very useful piece of information about an aquarium and one likely to be re-derived badly under pressure.
What comes out of the filter, and what stays in
Carbon out. Activated carbon exists to adsorb dissolved organic compounds from the water, and most medications are dissolved organic compounds. Leave it in and the treatment is removed as fast as you add it — a common reason a course “does not work”. The same goes for other chemical media: resins, phosphate removers, and anything a manufacturer describes as polishing the water. The filter media guide covers which cartridge contains what, and the answer is often “carbon, hidden inside the sponge”.
Biological media in, pump running. The sponges and ceramic where the nitrifying bacteria live must stay exactly where they are, and the flow must continue. Switching the filter off for a course of treatment starves the colony of oxygen and flow, and the longer it stays off the more of it dies, so you can emerge from the course with a tank that is no longer cycled. The filtration guide explains which part is which.
UV off. A UV clarifier or steriliser will degrade many medications as the water passes through it. Turn it off for the duration and back on afterwards.
Ozone and protein skimmers off, if you have them, for the same reason.
Oxygen, which is the part people skip
Increase aeration for the whole course, every time, whatever you are using.
Several things are working against the fish at once. Some treatments — formalin most notably — consume oxygen directly. Some affect gill function, which is exactly the tissue a sick fish can least afford to have compromised. Many protocols involve raising the temperature, and warmer water holds less oxygen. And a fish under treatment is stressed, which raises its demand.
So run an air stone, drop the water level slightly to increase surface agitation, or angle the filter return to break the surface. The aeration guide covers the mechanics. If the fish is gasping at the surface during treatment, aerate hard and test the water, and if it has not eased within an hour or two, suspect the illness or the treatment and follow the abort steps below.
Where to treat: the tank or a hospital tank
Both are legitimate, and the choice follows from what you are treating.
Treat the display tank when the organism lives in it. White spot is the clear case: the parasite spends part of its life cycle off the fish, in the substrate and the water, and treating the fish alone in a separate tank leaves the reservoir behind for it to return to. Anything affecting several fish is usually a tank problem too.
Use a hospital tank when one fish has a localised problem. An ulcer, a fungal patch, an injury. A bare tank gives you precise dosing with no substrate or decor absorbing the medication, keeps the treatment away from the display tank’s bacterial colony, and lets you observe the fish properly. It is the same equipment as a quarantine tank run a different way, which setting up a hospital tank goes through.
The cost is the move itself, which is a stress on an already unwell animal — handling, netting and moving a goldfish covers how to do it with the least damage — and the fact that a bare hospital tank has no established filter. Setting up a hospital tank covers it in full: a sponge filter seeded from the main tank, daily water changes to keep ammonia down, and otherwise kept simple — bare bottom, somewhere to hide, a lid, an air stone, a heater if the treatment calls for one.
| Treating in a | Display tank | Hospital tank |
|---|---|---|
| Breaks a parasite’s life cycle | Yes | Not by itself — reinfection on return, unless the tank is treated or left fishless |
| Dosing accuracy | Approximate; substrate and decor absorb | Precise |
| Risk to the biological filter | Real, with antibacterials | None to the main system |
| Stress on the fish | None beyond the treatment | The move itself |
| Cost of a mistake | Every fish in the tank | One fish |
Medication and your filter bacteria
Antibacterial treatments do not distinguish between the bacteria causing the problem and the bacteria running the nitrogen cycle. Some are harder on the filter than others, and a few — methylene blue is the usual example — are essentially incompatible with a biological filter, which is precisely why they are used in bare hospital tanks.
So if you treat a display tank with anything antibacterial, plan for it:
- Test ammonia and nitrite daily, during the course and for a week after.
- Be ready to do water changes on the strength of those readings, accepting that a change during treatment dilutes the medication and may mean re-dosing the fraction removed — read the product’s instructions on this point specifically.
- Do not also clean the filter in the same fortnight. The colony is under enough pressure.
- If the tank was recently set up, understand that you may be pushing an incomplete cycle back to the beginning.
The main classes, and what they are for
Without dosing figures, because those belong to the specific product in your hand and nowhere else.
Antiparasitics. The white-spot and general products are usually built around formalin, malachite green, or a combination of the two, and aimed at the external parasites: white spot and the organisms behind flashing and rubbing. For gill and skin flukes, veterinary sources name praziquantel as the drug of choice. They are effective and they are the class with the least margin for error. Malachite green stains silicone and decor, degrades in light, and is an irritant to handle — wear gloves. Formalin consumes oxygen. Both assume an accurate volume.
Copper-based treatments. Effective against several parasites and genuinely difficult to use well, because copper binds to substrate, rock and decor and then leaches back, so the concentration in the water is not the concentration you dosed. It is lethal to snails, shrimp and other invertebrates at treatment levels. In a bare hospital tank with a test kit it is a tool; in a planted display tank with gravel it is a way to produce an uncontrolled exposure.
Antibacterials and antiseptics. The over-the-counter products are mostly surface-acting and are reasonable for external bacterial problems — the early stage of an ulcer, fin damage that is not healing. They do little against an infection that has gone internal, which is what red streaks and a fish that is systemically unwell usually represent.
True antibiotics. In the UK and the EU these are prescription-only for fish by law, as they are for other animals. In the US none is approved for aquarium fish, and those sold in shops and online are unapproved and illegal to market. Either way the rules are not a formality — the drug that works depends on the organism, and the choice is a clinical one. A fish with a systemic infection needs a vet who treats aquatic patients, and antibiotic resistance is a real reason the rules exist. Over-the-counter products that imply antibiotic effect without a prescription generally are not one.
Antifungals. Often the same active ingredients as the antiparasitics. Genuine fungal growth in goldfish is often secondary to an injury or bad water, so the fungus page starts with the wound rather than the mould, and so should you.
Salt. A legitimate treatment with a genuine evidence base, and the one most often used without thinking of it as a medication. It has its own page — aquarium salt for goldfish — and the same rules apply: dose on real volume, add it dissolved rather than as crystals, replace only the fraction removed by a water change, and do not stack it on top of another treatment without knowing that the two are compatible.
What does not work
Treating without a diagnosis. Putting a general-purpose product in because a fish looks off is how tanks end up with three treatments in a fortnight and a filter that has stopped working. If you cannot name a sign, go back to testing the water and watching.
Multi-purpose “cure everything” products. A single bottle claiming to address parasites, bacteria, fungus and stress is either a mild antiseptic that does little, or a collection of active ingredients none of which is at a useful concentration.
Tea-tree and herbal extracts. The evidence that they do anything therapeutic is thin. They also tend to form an oily film on the surface, which reduces gas exchange at exactly the moment a sick fish needs oxygen most. If you use one, run an air stone and skim the surface.
Preventative dosing. Medicating a healthy tank on a schedule achieves nothing, stresses the fish, selects for resistant organisms and periodically damages the filter. Prevention in fishkeeping is water changes, appropriate stocking and quarantine — not chemistry.
Stopping early. A fish looking better on day three of a seven-day course is a fish whose parasites have been knocked back rather than eliminated. Finish the course as written.
Temperature, and the season you are in
Most treatments assume a metabolism that is running: the fish’s immune system does part of the work, and the parasite’s life cycle has to reach the stage the medication can act on. Both slow as the water cools.
In practice that means a problem in a pond in early spring is harder to deal with than the same problem in July — the course takes longer, the medication does less, and the fish contributes less. It is one of the arguments for dealing with a sick fish in autumn rather than hoping it holds until spring, which the seasonal feeding page makes from the other direction, and one of the arguments for bringing a sick pond fish indoors where the temperature can be controlled. The temperature guide covers what a goldfish tolerates and how quickly it can be changed, which is slowly.
During and after
Feed lightly or not at all. A fish under treatment is often not eating properly, and uneaten food in a tank you cannot clean normally makes the water worse.
Do not vacuum the substrate mid-course unless the instructions say to, and be aware that any water you remove takes medication with it.
Look at the fish rather than the calendar. The healthy baseline page describes what improvement looks like: fins carried open again, normal colour returning, interest in food, normal position in the water.
When the course is finished: a large water change, fresh activated carbon back in for a few days to strip the residue (after copper, for at least a week after a copper test reads zero), the UV back on, and ammonia and nitrite tested daily for a week. Then leave the tank alone.
A safety note about the medications themselves
They are not harmless to you either. Several are irritants, malachite green and formalin in particular, and formalin fumes are unpleasant and should not be inhaled over a bucket. Wear gloves, keep the lid on, and do not measure them out over food preparation surfaces. Store them out of reach of children and animals, keep them in the original container with the instructions attached, and dispose of unused product as household chemical waste rather than down a drain that reaches a watercourse.
The five checks before you dose
| Check | Why |
|---|---|
| Have I tested the water? | Most problems are environmental and medication will not fix them |
| Do I know my real volume? | The nominal figure overdoses by 10–25 percent |
| Is the carbon out and the UV off? | Either will remove the treatment |
| Is aeration increased? | Several treatments reduce available oxygen |
| Is anything else in the water? | Do not run two treatments together unless the label or a vet says they can be combined |
If all five have an answer, dose properly, finish the course, and watch the fish. If any of them does not, that is the thing to do next — not the medication.
Frequently asked questions
How do I work out the dose for my tank?
On the volume of water actually in it, which is always less than the tank's nominal size. A tank sold as 100 litres is filled to a few centimetres below the rim, and gravel, rocks, decor and equipment displace more, so the real figure is commonly 10 to 25 percent lower. Take the internal length, width and filled depth in centimetres, multiply them, divide by 1,000 for litres, then subtract a sensible allowance for everything sitting in it. Dosing a nominal volume is a common way people overdose a goldfish.
Do I need to take the filter out while treating?
No — take the carbon out and leave the filter running. Activated carbon and other chemical media are designed to adsorb dissolved organics and will strip the medication out of the water, which is why a treatment sometimes appears to do nothing. The biological media, the sponges and ceramic where the bacteria live, must stay in and the pump must keep running: switching the filter off for a course of treatment starves the colony of oxygen and flow, and the longer it stays off the more of it dies, so you can end up with an ammonia problem on top of a sick fish.
Can I use two medications at once to be thorough?
Not unless the label or a vet says the two can be combined. Combinations are unpredictable, the safety margin of each product assumes it is the only thing in the water, and a fish already weakened by illness is the worst possible subject for an experiment. If one treatment has not worked, stop it, do a large water change, run fresh carbon for a day or two to strip the residue (after copper, keep it in for at least a week after a copper test reads zero), and only then start something else. The one exception is a commercial product that already contains two active ingredients, which has been formulated and tested as a combination.
Is it safer to use half a dose?
No, and it is a surprisingly common way to lose a fish. An underdose gives the parasites or bacteria an exposure they survive, which selects for the ones least affected and leaves the fish still infected after a full course of stress. If the correct dose worries you, the answer is to measure the volume more carefully rather than to guess low. The exception is a product that explicitly gives a reduced dose for sensitive species or for scaleless fish — goldfish are not in that category.
Should I treat the main tank or move the fish?
It depends what you are treating. A parasite with a free-swimming stage, white spot being the obvious case, lives in the tank as well as on the fish, so the display tank has to be dealt with too, by treating it or by leaving it fishless for several life cycles of the parasite. A single fish with an ulcer or a fungal patch is often better moved to a bare hospital tank, where dosing is precise, there is no substrate soaking up the medication, and the display tank's filter is left alone. Moving a fish is itself a stress, so it is a judgement rather than a rule.
My fish looks worse an hour after dosing. What do I do?
Abort. Do a large water change immediately, 50 percent or more, add fresh activated carbon to strip what is left, and increase aeration. Gasping, rolling, darting, lying on its side or a sudden loss of colour shortly after a dose means the fish is not tolerating it, and no treatment is worth pursuing through that. Work out afterwards whether the dose was miscalculated, whether the volume was wrong, or whether the product is simply not tolerated.
Are the antibiotics sold for aquariums real antibiotics?
Often not. In the United Kingdom and the European Union, antibiotics for fish are prescription-only by law, so what is sold over the counter is usually an antibacterial or antiseptic of a different kind — useful for surface problems, largely useless against an internal infection. In the United States no antibiotic is approved for aquarium fish, and the FDA treats the ones sold in shops and online as illegal to market, so a vet's prescription is the legitimate route. If a fish has a systemic bacterial infection, the treatment that works is a prescription one, which means a vet who sees aquatic patients. Products promising antibiotic results without a prescription are worth reading sceptically.
Does temperature affect how well a treatment works?
Considerably. Most treatments assume a fish whose metabolism is running and an immune system doing part of the work, and both slow as the water cools. In cold water a parasite's life cycle also slows, which lengthens the course needed. This is one reason a problem in a pond in early spring is harder to deal with than the same problem in summer, and one of the arguments for treating a fish indoors rather than waiting for the weather.