The primary goal of this research was to predict how changes in water temperature affect the swimming performance and energetic cost of transport in adult pink salmon (Oncorhynchus gorbuscha) and sockeye salmon (O. nerka), thus contributing to their ability to reach natal streams and spawn successfully. The prolonged swimming performance (U[sub crit]), minimum and maximum metabolic rate (Mo[sub 2-min] and Mo[sub 2-max]), oxygen cost of transport (COT) for upper Fraser River pink salmon were assessed across a range of naturally occurring temperatures using Brett-type swim tunnel respirometers and compared with values for sockeye salmon. To reduce mortality in senescing fish we minimized holding time and, therefore, thermal acclimation to as little as 48 hours before experiments. Therefore, we also used a salmonid model, the cutthroat trout (O. clarki clarki), to examine the effects of 48-hour and 3-week temperature acclimation periods on U[sub crit]. The length of the acclimation period had no significant effect on either the first or second U[sub crit] or on the recovery ratio (the quotient of U[sub crit-2]/U[sub Crit-1])- These results indicate that a 48-h acclimation to experimental temperatures may be sufficient in studies of swimming performance with this species. Contrary to previous beliefs, pink salmon were capable of similar relative critical swimming speeds as sockeye salmon (2.25 FL•s⁻¹), but sockeye salmon swam to a higher absolute U[sub crit](125.9 cm•s⁻¹) than pink salmon (116.4 cm•s⁻¹) because of their larger size. However, some individual pink salmon swam faster than all the sockeye salmon tested. Metabolic rate increased exponentially with swimming speed (P < 0.01) in both species and was higher for pink than sockeye salmon (P = 0.01), although swimming efficiency (was not significantly different between the species at their optimal swimming speeds minimum cost of transport; COT[sub min]) . The upper and lower limits of metabolism also increased exponentially with temperature (Mo[sub 2-min] P = 0.01; Mo[sub 2-max] , P < 0.01, respectively) but were not different between species (Mo[sub 2-min] P = 0.93; Mo[sub 2-max] , P = 0.38). The relationship between Mo₂ and swimming speed was positively affected by temperature in pink salmon (P = 0.01), but average and minimum COTs were independent of temperature over the range tested (9-22 ºC) in both species. Overall, a higher degree of inter-individual variability and thermal insensitivity in pink salmon suggest that this species might not be as locally adapted to particular upriver migrations as are sockeye salmon.