Six paired lots of yearling lake trout (Salvelinus namaycush) reared at densities of 41,000 and 51,000 fish per raceway during their last 9 months in the hatchery were stocked in Lake Ontario. Poststocking survival of the high-density (HD) and low-density (LD) fish was not different for the 1982 year-class. However, for the 1983 year-class, mean survival was significantly different between HD and LD fish (P < 0.01). Mean survival of HD fish was only 76% that of LD fish (P < 0.01), and most of the mortality attributable to rearing conditions had apparently occurred within 2 months after stocking. Mean size at stocking was not different for HD and LD fish of the 1982 year-class, but for the 1983 year-class, the LD fish were 6% longer and 22% heavier than the HD fish. Mean lengths and weights of LD and HD fish were not different in samples collected in Lake Ontario at age 2 and older. Size at stocking was not likely the factor that caused the difference in survival. Rather, the rearing conditions (probably water exchange rate in relation to number of fish in the raceway) that resulted in slower growth of the HD fish of the 1983 year-class also caused them to be poorer physiologically than the LD fish. The number of yearling lake trout per rearing unit that will result in maximum contribution to populations in the Great Lakes after stocking may be lower than the rearing densities customarily used at some hatcheries.