Effects of transmitter attachment procedures on smolts of Atlantic salmon Salmo salar have primarily been examined in the laboratory with hatchery-reared fish. Biotelemetry researchers must extrapolate findings of tagging studies to wild, migrating smolts. Because hatchery rearing can affect salmonid development and because field tagging procedures and environmental conditions can differ from those in the laboratory, we measured critical swimming speed of hatchery-reared Atlantic salmon smolts equipped with internal, external, or gastric dummy radio transmitters in the laboratory. We then repeated the experiment with wild, migrating fish in ambient environmental conditions. Swimming ability of hatchery smolts was not significantly affected by any attachment procedure, and no significant differences were found between fish tested 1 or 16 h after tagging. Similarly, wild smolts were not impaired by gastric transmitters; however, performance of wild fish tested 1 or 16 h after internal or external attachment was significantly lower than that of wild controls. Thus, reaction of hatchery-reared Atlantic salmon smolts to attachment procedures can differ from that of wild migrants. Biotelemetry researchers should exercise caution in interpreting data obtained from wild Atlantic salmon smolts equipped with internal or external radio transmitters during the first 16 h after tagging.