During peak smoltification, the interactive effects of cortisol and ambient seawater challenge were compared in two strains of Atlantic salmon (Salmo salar) smolts: a domesticated strain, AquaGen and a native River Imsa strain. Tissue and blood samples were taken from untreated fish on 20 May. Fish were then transferred to experimental tanks, allowed to recover for 24 h and cortisol (50 mg kg body mass-1), dissolved in vegetable oil, or vegetable oil alone (sham) was implanted. Samples were taken 5 days post implantation. Fish were then exposed to 24 h ambient seawater challenge (FW-SW) or freshwater to freshwater replacement (FW-FW) and sampled as before. Sham implantation had no significant impact on any of the measured parameters. Cortisol implantation significantly elevated plasma cortisol in FW-FW Imsa smolts, while no effect was observed in FW-SW Imsa smolts. Cortisol implantation had no effect on the plasma cortisol levels of AquaGen smolts regardless of FW-SW challenge. Increased plasma cortisol corresponded with significantly higher plasma glucose levels in FW-FW Imsa smolts. Plasma Na+ and Cl- levels were not affected by cortisol implantation but were significantly increased in FW-SW smolts of both strains. Gill Na+,K+-ATPase activity increased in response to cortisol implantation in only FW-FW smolts but not FW-SW smolts. Gill Na+,K+-ATPase ??-subunit mRNA levels were not affected by strain, cortisol injection or transfer protocol, while both CFTR I and CFTR II mRNA levels were significantly higher in AquaGen versus Imsa smolts regardless of treatment. CFTR I mRNA was elevated following cortisol implantation in FW-FW smolts from both strains suggesting CFTR I expression is under the control of cortisol. These findings also suggest that plasma cortisol levels are regulated differently between strains and that cortisol implantation and ambient FW-SW challenge interact, interfering with the individual effects of each of these factors. ?? 2003 Published by Elsevier Science B.V.