Survival of the 1963–1982 year-classes of stocked yearling lake troutSalvelinus namaycushdeclined significantly over time in Lake Superior. To investigate possible causes of this decline, a Ricker model of stock–recruitment was used to describe the catch per effort (CPE) of age-7 stocked lake trout in the Michigan, Minnesota, and Wisconsin waters of Lake Superior as functions of the numbers of yearlings stocked 6 years earlier (an index of density dependence), the density (CPE) of wild adult lake trout (an index of predation), and large-mesh (≥114-mm stretch-measure) gill-net fishing effort (an index of fishing mortality). Declining CPE of stocked lake trout in Michigan and Wisconsin was significantly associated with increasing large-mesh gill-net fishing effort. Declining CPE of stocked lake trout in Minnesota was significantly associated with increasing density of wild lake trout. Declining survival of stocked lake trout may therefore have been caused by increased mortality in large-mesh gill-net fisheries in Michigan and Wisconsin, and by predation by wild lake trout that recently recolonized the Minnesota area. We recommend that experimental management be pursued to determine the relative importance of large-mesh gill-net fishing effort and of predation by wild lake trout on the survival of stocked lake trout in U.S. waters of Lake Superior.