The FDA's safe minimum internal temperature for seafood is 145°F, and that number applies across the board: salmon, cod, tilapia, mahi, shrimp, scallops, and shellfish generally. What changes by type isn't the safety threshold, it's the texture people actually prefer, which is why a salmon fillet and a scallop don't always get cooked to the exact same finishing point in practice.
Knowing the difference between the safety number and the texture preference is the useful part. One keeps you safe. The other is about how the fish actually tastes once it's on the plate.
The baseline: 145°F applies to all seafood
FDA guidance sets 145°F, held for 15 seconds, as the safe minimum internal temperature for seafood generally. That includes finfish like salmon, cod, tilapia, and mahi, and it includes shellfish like shrimp, scallops, and lobster. There isn't a separate, lower official number for any specific type; the destination is the same, even though how fast different seafood gets there varies quite a bit by thickness and fat content.
This baseline exists because the same underlying risk, mainly Vibrio and related bacteria that occur naturally in marine environments, is a concern across seafood broadly, not just one category of it. Cooking to temperature is what addresses that risk, regardless of which type of seafood is on the cutting board.
Where texture preferences diverge from the safety number
Once you're past the safety threshold, texture preference is where seafood starts to differ by type, and this is where a lot of the confusion about "what temperature" comes from.
Salmon and other fatty fish. Many cooks pull salmon closer to 125 to 130°F, well under the 145°F mark, aiming for a moister, more medium result rather than the firmer, more well-done texture that 145°F produces. This is a texture choice, not a food-safety recommendation, and it's one that assumes fresh, properly handled fish. Cooking salmon fully to 145°F is still the safe, standard approach, and it's the one to default to unless you specifically know why you're deviating from it.
White, leaner fish. Cod, tilapia, flounder, and mahi have less fat than salmon, and they tend to dry out faster once they pass done. These are generally cooked closer to the full 145°F standard, since there's less margin for a lower finishing temperature before the texture suffers rather than improves.
Scallops. Scallops are often cooked to around 130°F internally, pulled once the flesh turns milky white and opaque and feels firm but still yields slightly. Past that point, scallops toughen quickly, similar to how shrimp and lobster behave when they're pushed past done.
Shrimp and lobster. Both are generally cooked to the full 145°F standard, since a firmer, opaque result is the texture goal for both, not a moister undercooked one. If you want the visual and textural cues specific to each, the same logic covered in how to check shrimp for doneness and how lobster tail signals it's done applies here too: opaque color, firm texture, and a curled or contracted shape are the shared signs across most seafood.
The rare tuna exception, and why it's not a loophole
Seared rare tuna is the seafood world's clearest exception to 145°F, and it's worth understanding why it's treated differently instead of assuming it means the general rule is optional.
Rare or seared tuna relies on sushi-grade fish that's been flash-frozen to specific FDA parasite-destruction standards, sourced and handled under conditions meant to control the specific risks that come with serving fish well under the standard cooking temperature. Restaurants serving it this way are expected to carry a consumer advisory noting the risk of consuming undercooked seafood. That combination, specific sourcing plus freezing plus disclosure, is what makes rare tuna a controlled exception rather than a general "some fish don't need to be cooked through" rule. It doesn't extend to an ordinary fillet from the grocery store handled with typical fresh-fish precautions, and it isn't a template for cooking other seafood types below temperature.
Checking temperature correctly
A calibrated instant-read thermometer inserted into the thickest part of the fish or shellfish is the most reliable check across every type covered here. For a fillet, that's usually the center, away from thin edges that cook faster and will read a misleadingly high temperature. For shellfish like shrimp or scallops, the center of the largest piece in a batch is the one to check, since smaller pieces in the same pan will finish first.
Visual and textural cues back up a thermometer reading and are useful when you don't have one in hand. Across nearly all seafood, doneness follows a similar pattern: translucent, slightly gray or glassy flesh turns opaque and takes on its cooked color, whether that's white, pink, or a mix depending on the type, and texture firms up from soft to a slight spring, without going all the way to tough.
A thermometer's accuracy matters more than people usually give it credit for. A cheap or uncalibrated thermometer can read several degrees off, which is enough to make the difference between correctly reading 145°F and pulling fish that's actually still a few degrees under. Checking a thermometer's accuracy against boiling water, which should read 212°F at sea level, is a quick way to know whether the reading you're trusting is actually right.
How thickness, not weight, determines cook time
Two fillets of the same weight can cook at very different speeds if one is thick and compact and the other is thin and wide. Cooking time scales with thickness far more than with total weight, which is why a recipe that gives a cook time based on weight alone is only a rough guide, not a guarantee.
A rule some cooks use as a starting point is roughly 10 minutes of cooking time per inch of thickness at moderate heat, checked with a thermometer near the end rather than relied on exactly. This works reasonably well for finfish cooked whole or as fillets, though it's less useful for shellfish like shrimp and scallops, which cook fast enough that a minutes-per-inch rule doesn't leave much room for the check to matter. For anything with meaningfully uneven thickness, like a fillet that's thick at one end and thin at the other, checking the thick end and expecting the thin end to finish a little earlier is more useful than averaging the two.
Storage and thawing affect how evenly seafood cooks
Frozen seafood that's thawed unevenly, or cooked straight from frozen without adjusting, is one of the more common reasons a piece of fish or shellfish comes out overcooked on the outside and underdone in the center. Thawing in the refrigerator overnight, rather than at room temperature or in a hurry under warm water, gives the most even starting point and the most predictable cook time once it hits the heat.
Seafood that's still cold in the center when it goes into the pan will need more time than the same seafood at room temperature, and this is one of the more common reasons a fillet reads unevenly on a thermometer: the outside is already past 145°F while the center, cooking from a colder starting point, is still catching up. Letting fish sit out briefly before cooking, within food-safety limits for how long raw seafood can safely sit at room temperature, closes that gap somewhat and makes for a more even result.
Carryover cooking applies to seafood too
Seafood keeps cooking for a short time after it's removed from heat, the same carryover effect that happens with meat and poultry, though it's smaller in magnitude since most seafood cuts are thinner and hold less residual heat overall. This matters most for thicker cuts, a full salmon fillet or a larger scallop, and matters less for something thin like a shrimp or a thin fish fillet, where there's simply less mass to hold extra heat.
Pulling a thicker piece of fish a few degrees before its final target, then letting it rest briefly, often lands closer to the actual target than cooking it the entire way on direct heat and pulling exactly at temperature, since the extra couple of degrees from carryover cooking would otherwise push it past done. This is more of a texture refinement for thicker cuts than something that changes the safety threshold itself, and it becomes less relevant the thinner or smaller the piece is.
Why this varies by menu item, even at the same restaurant
If you've ordered fish at a seafood restaurant and noticed one dish comes out flakier and another firmer, that's usually the type-specific texture target at work, not inconsistency in the kitchen. Atlantic Salmon and Lemon Basil Mahi are both real finfish entrees that get handled a little differently for exactly this reason, salmon's higher fat content gives it more room for a slightly lower finishing point, while leaner white fish like mahi is typically cooked closer to done throughout. Bacon-Wrapped Sea Scallops, listed on the full dinner and starters menu, is a good example of the scallop-specific target described above: opaque, firm, and pulled before the texture turns rubbery.
The takeaway that applies across all of it: 145°F is the number to know for safety, and it's the right target by default for shrimp, lobster, scallops, and lean white fish. Fattier fish like salmon has more legitimate room for a lower finishing temperature if that's the texture you're after, and rare tuna is a specific, controlled exception with its own sourcing requirements, not a general shortcut. When in doubt, especially for anyone in a higher-risk group, 145°F held for 15 seconds is the number that doesn't require any caveats.