Swimming stamina, buoyancy, and gill Na +K + ATPase activity were measured in yearling spring chinook salmon before and after release from Dworshak National Fish Hatchery and Rapid River Hatchery from 1986 to 1988. Behavioral and physiological indices remained constant during the 3 months prior to release in early April. Following release, freeze-branded smolts were sampled at Lower Granite Dam, 118 and 279 km downstream from the respective hatcheries. These smolts exhibited swimming stamina and gill Na +K + ATPase levels that were significantly higher than those measured at the time of release. Buoyancy decreased among the same fish (significantly for one group). Migration rates to Lower Granite Dam were slow, but increased as smolts migrated further downstream. Experimental manipulation of water temperatures in the laboratory, using Little White Salmon National Fish Hatchery stock, produced smolts with behavioral and physiological responses similar in magnitude to those measured in actively migrating smolts from the Idaho hatcheries. The commonly held premise that, as a result of smoltification, migrating fish exhibit decreased swimming stamina and increased buoyancy during migration, is not supported by this study. Furthermore, the results indicate that the general hatchery population of yearling chinook salmon received insufficient environmental cues necessary for full smolt development, prior to release.