We tested the null hypothesis that differences in the seasonal return patterns between stocks of Atlantic salmon, Salmo salar L., are a result of a direct response to the environment, and not under genetic control. Two stocks were used in the experiments, originating from the R. Figga and R. Imsa, respectively. In their native habitat fish from the former are known to return to the home stream as adult salmon early in the summer, while from the latter return during late summer and autumn. By rearing these stocks in the same hatchery and releasing smolts of both stocks together at three sites in southern Norway, it was demonstrated that salmon from the R. Figga stock returned earlier to coastal Norway than salmon from the R. Imsa stock, as maturing adults. Thus, we reject the hypothesis that these stocks are genetically identical in this trait. Within both stocks, multi-sea-winter fish returned earlier than one-sea-winter fish. Within stocks, there was no significant difference in time of return between salmon released as 1- and 2-year-old smolts, or between fish reared from parents ascending the R. Imsa early or late in the season.