How Long to Smoke Salmon: One Name, Two Different Fish

photo · how-long-to-smoke-salmon-one-name-two-different-fish

Traeger says 180°F for 3 to 4 hours. Extraordinary BBQ says 250°F for 45 minutes to an hour. A Reddit thread that Google’s own AI Overview cites as a source says 225°F for about 90 minutes. All three are answering the same search, and the numbers disagree by a factor of four. Then the AI Overview does something it does not do on any other cut covered on this site: instead of picking one of those numbers, it stops and asks a question back. “Are you hot-smoking or cold-smoking? Did you dry brine or wet brine the fish?”

That is not the model hedging. It is a real fork in the question. This site already ran into one word covering two different foods on ham, where a fully-cooked deli product and a raw roast share nothing but a name. Salmon splits the same way, and the two versions sit further apart than a spiral-cut ham and a fresh one. Hot-smoked salmon is cooked through, flaky, and safe to eat straight off the smoker. Cold-smoked salmon, the lox-style product sliced thin at a deli counter, never gets hot enough to cook at all and depends on salt and time rather than heat to make it safe. Everything below is the hot-smoked version. It is the one nearly every recipe in the search results actually means, and the one most people typing this question are asking about.

A fish with nothing to convert

Every cut covered on this site so far has had some amount of collagen setting its clock: a lot on a pork shoulder, less on a chuck roast, barely any at a chicken wing joint, none at all on a turkey breast. Salmon sits past the far end of that list. It never carried weight the way a shoulder or a thigh did. A fish’s muscle is built from short blocks called myomeres, stacked and separated by thin sheets of connective tissue, not the long working fibers that make up a cut of beef or pork. There is no tough tissue waiting to break down into something better, and no stall to wait out the way there is on a brisket. The fish is done once heat has moved from the surface to the center. Every recipe in the search results is really describing that same diffusion problem at a different chamber temperature, not a different kind of cook.

Time by chamber temperature

None of these rows are wrong. They are the same diffusion measured at different rates, the way water boils faster on high heat than on low. What decides which row actually applies to a given fillet is not which recipe someone followed, it is how thick the fish was to start with, the same variable that already separates a quick wood chip burn from a slow split log on this site’s fuel pages, and the same reason a whole chuck roast takes four times longer than the same meat cubed. A thin tail piece and a thick center-cut fillet are not the same cook even in the same smoker at the same setting, and several of the times cited above almost certainly come from filets of different thickness without saying so.

Why a fish farm’s numbers look different from a backyard recipe’s

The two-stage method in that table, starting at 140-150°F for about an hour before raising the chamber to 175°F, comes from Mowi, one of the world’s largest salmon producers, not from a home cook. That is worth noticing. A backyard recipe only has to work once, for one fillet, on one afternoon, so it optimizes for a single clear number and a predictable finish time. A producer is running the same process across thousands of fillets that all need to come out with the same moisture and the same texture, batch after batch, and a slower ramp gives more room to catch a fillet that is drying out before it overshoots. The destination temperature is close to what the hobbyist recipes land on. The path there is more cautious, because the cost of a bad batch is completely different at the two scales.

The pellicle is doing the work the fire gets credit for

Nearly every recipe in the search results includes a step that has nothing to do with the smoker: a salt brine or dry cure, followed by an hour or more of air-drying in the fridge, uncovered, before the fish goes anywhere near smoke. That step builds what is called a pellicle, a thin, tacky layer that forms on the surface as moisture is pulled out and redeposited. Smoke does not bond well to a wet surface. It bonds to that tacky layer. Skip the cure and the air-dry, and a fillet can sit in a smoker for the full time listed in the table above and come out pale and bland, because the smoke had nothing to stick to. Wood choice matters less than most guides suggest, this site’s own wood guide for salmon makes that case already, but skipping the pellicle undercuts every wood choice equally. Every time in that table assumes the pellicle is already there.

Why the safe number is so much lower than everything else on this site

Pork, beef, and chicken on this site all carry safety floors between 145°F and 165°F. Salmon’s hot-smoked target sits at 135-140°F for a flaky texture, or 145°F for a firmer, more conservative result, well under what would be considered done on any cut of meat covered so far. That gap is not an inconsistency. Fish muscle cooks through faster and firmer at a lower temperature than red meat or poultry, because myomeres are short and separated by connective sheets that soften early, unlike the long fibers in a piece of beef. The white protein that beads up on an overcooked fillet, called albumin, starts coagulating and pushing to the surface well before 140°F, which is also why pulling salmon at the low end of its range keeps it moist, while chasing a number closer to chicken’s 165°F floor turns it chalky and dry with nothing gained in return.

What we have not measured

We have not run a thin tail cut against a thick center fillet side by side at the same chamber temperature to log the actual time gap between them. The diffusion argument above follows from how heat moves through food generally, and from what this site has already worked through on wood and on cuts with far more mass, not from a dedicated salmon trial. A sausage’s diameter is fixed by the casing rather than chosen by the cook, and a salmon fillet’s thickness is fixed by the fish it came from in much the same way, but neither substitutes for a thermometer. Treat the times in the table as a starting point tied to a reading at the thickest part of the fillet, not a replacement for one. See how we test for what “measured” means on this site.

The pellicle argument here assumes a fillet, which is one large face of exposed flesh. It does not survive the move to a fish sold whole. Smoked trout arrives wrapped in skin, so most of its surface cannot form a pellicle at all, and the timing question turns into a question about how the fish was cut open. On smoked catfish, the same pellicle logic reappears as a full salt-and-sugar cure that splits the recipes into two different clocks.