We compared the relative abundance of lake trout Salvelinus namaycush spawners in gill nets during fall 1999–2001 in Lake Michigan at 19 stocked spawning sites with that at 25 unstocked sites to evaluate how effective site-specific stocking was in recolonizing historically important spawning reefs. The abundance of adult fish was higher at stocked onshore and offshore sites than at unstocked sites. This suggests that site-specific stocking is more effective at establishing spawning aggregations than relying on the ability of hatchery-reared lake trout to find spawning reefs, especially those offshore. Spawner densities were generally too low and too young at most sites to expect significant natural reproduction. However, densities were sufficiently high at some sites for reproduction to occur and therefore the lack of recruitment was attributable to other factors. Less than 3% of all spawners could have been wild fish, which indicates that little natural reproduction occurred in past years. Wounding by sea lamprey Petromyzon marinus was generally lower for Seneca Lake strain fish and highest for strains from Lake Superior. Fish captured at offshore sites in southern Lake Michigan had the lowest probability of wounding, while fish at onshore sites in northern Lake Michigan had the highest probability. The relative survival of the Seneca Lake strain was higher than that of the Lewis Lake or the Marquette strains for the older year-classes examined. Survival differences among strains were less evident for younger year-classes. Recaptures of coded-wire-tagged fish of five strains indicated that most fish returned to their stocking site or to a nearby site and that dispersal from stocking sites during spawning was about 100 km. Restoration strategies should rely on site-specific stocking of lake trout strains with good survival at selected historically important offshore spawning sites to increase egg deposition and the probability of natural reproduction in Lake Michigan.