There is a sentence in the FDA-based guidance for fish processors that settles most of the argument about dried and salted fish. Any product preserved by salting, drying, pickling or fermentation must be eviscerated before processing, because without evisceration, toxin formation is possible during the process even with strict control of temperature.
Not “should be.” Not “is safer if.” The requirement exists because the two things people rely on at home, salt and refrigeration, both fail in specific and well-documented ways when the fish still has its insides.
For scale, Ukraine’s Public Health Center recorded 46 botulism clusters in the first half of 2025, with 57 people affected, seven of them children. Forty-two diagnoses were confirmed, three people died, and 45 patients had to be given antitoxin. Twelve more cases were registered across six regions in the first four months of 2026.
The spores live in the guts
Clostridium botulinum is ordinary in the environment. In fish, its spores turn up in the gills and viscera, and type E, the strain behind fish-borne botulism, is the form most commonly found in fresh water and marine environments. The guidance puts it bluntly: assume C. botulinum is present in any raw fishery product, particularly in the viscera.
Leaving the viscera in place does two things at once. It keeps the spores inside the fish, and it builds the anaerobic pocket they need. As the fish dries, the space under the skin and along the backbone runs out of oxygen. That is the environment type E wants, and salt reaches it last.
The guidance also closes the obvious loophole: evisceration has to be thorough, and if even a portion of the viscera or its contents is left behind, the risk of toxin formation remains.
What “well salted” actually means
The everyday assumption is that a lot of salt means everything is dead. Food microbiology measures something different: water phase salt, the share of salt relative to salt plus moisture inside the flesh, not the salt sitting on the skin.
The calculation is percent salt divided by percent salt plus percent moisture, times one hundred. The catch is that both numbers have to be measured in an actual piece of fish, and the piece that matters is the one likely to have the lowest water phase salt, which is the thickest part near the spine.
Here are the benchmarks industry works to.
| Condition | Water phase salt |
|---|---|
| Small fish under 5 inches, exempt from the evisceration requirement | 10% in refrigerated products |
| Shelf-stable caviar, New York requirement | at least 9% |
| Product stored at 38°F minimum | under 5% |
| Product stored at 50°F minimum | 5% or greater |
Nine or ten percent water phase salt is a very salty product, and naming that figure requires a laboratory. Home salting offers no assurance that the concentration near the backbone gets there, and no way to check. Taste does not help, because the tongue reads the surface.
So salting does not fail because salt is powerless. It fails because home salting rarely drives salt to the required share in the place it is required, and gives no signal either way.
The refrigerator does not stop it either
The second assumption is that cold halts growth. For type E that is false.
The minimum temperature for growth and toxin formation by C. botulinum type E and nonproteolytic types B and F is 38°F, or 3.3°C. That is an ordinary refrigerator shelf. Type A and proteolytic types B and F need noticeably more warmth, which is why home canning and dried fish behave differently and need different rules.
Surveys of domestic refrigerators find temperatures above 50°F more often than anyone would like. Fish hanging in a kitchen or on a balcony is outside temperature control altogether.
Nothing about the fish looks or smells wrong
Fish carrying botulinum toxin does not swell, smell off or taste different. A canning jar at least sometimes bulges, and even that sign is unreliable. Dried fish offers no equivalent hint.
Toxin distribution within a single fish is also uneven. Several people can eat from the same carcass and not all of them get sick, which regularly misleads families into ruling the fish out.
What destroys the toxin and what does not
- The toxin is destroyed by heat. Boiling for 10 to 15 minutes inactivates it.
- The spores survive boiling. Killing them takes a pressure process above the boiling point.
- Drying, curing, smoking and salting are not thermal processing at all. They do not clear even the first bar.
Which leads somewhere uncomfortable: dried fish belongs to the category eaten with no step capable of neutralizing toxin. That is exactly why evisceration before salting sits in the rules as a requirement rather than advice.
Home canning carries its own separate rule set, and the reason cucumbers and mushrooms cannot be put up the same way is worked through in our piece on botulism in the jar.
Symptoms, and why hours matter
Onset is usually 12 to 36 hours after the meal, though it can arrive within hours or take several days. A later onset generally means a smaller dose.
Botulism descends, and that pattern is its signature. Cranial nerves go first: double vision, a drooping eyelid, dry mouth, altered voice, trouble swallowing. Weakness then moves down to the neck, the arms and the muscles of breathing, symmetrically on both sides.
One detail governs treatment. Antitoxin neutralizes only toxin still circulating in the blood. Whatever has already bound to nerve endings cannot be pulled back off, and recovery there runs for months as new endings grow. So the 2025 figure of 45 patients out of 57 receiving antitoxin matters alongside a question the statistics do not answer: on which day it was given.
Waiting until morning to see whether it passes works against the patient here, in the same way it does with mushroom poisoning, where feeling better is the dangerous phase. That scenario is covered separately in the hours-to-symptoms breakdown.
What to do in practice
- Buy only eviscerated dried and salted fish. A whole carcass with the guts in is the exact construction the rules forbid salting.
- Skip informal sellers. No process, no water phase salt measurement, no accountability.
- If you cure fish yourself, gut it thoroughly. Gut contents smeared on the flesh put the risk back.
- Do not lean on the refrigerator. Type E is satisfied with 3.3°C.
- Go to hospital on the first eye or swallowing symptom and name the fish, even if the meal was yesterday and everyone else is fine.
The bottom line
The hazard is not fish, it is one specific construction: a carcass with its viscera left in, dried without oxygen inside and without any heating. The spores are in the guts, the drying itself builds the anaerobic pocket, and salt arrives at the backbone last and at a share no kitchen can measure.
Cold is not a barrier either, since type E grows from 3.3°C. The toxin dies with 10 to 15 minutes of boiling, and nobody boils dried fish. Hence the rule stated without softening: without evisceration, toxin can form even under strict temperature control.
The practical version is short. Eviscerated fish from a known source, no whole carcasses from informal sellers, and a hospital the moment vision doubles or swallowing gets hard.
Cold fails as a barrier well beyond botulism. Listeria grows from 31.3°F upward and gives no sensory warning, while the average measured refrigerator sits at 42.3°F: what that means for ready-to-eat food.
