Information on predator and prey distributions is integral to our understanding of migratory connectivity, food web dynamics and ecosystem structure. In marine systems, although large animals that return to land can be fitted with tracking devices, minimum instrument sizes preclude deployments on small seabirds that may nevertheless be highly abundant and hence major consumers. An increasingly popular approach is to use N and C stable isotope analysis of feathers sampled at colonies to provide information on distribution and trophic level for the preceding, and generally little-known, nonbreeding period. Despite the burgeoning of this research, there have been few attempts to verify such relationships. In this study, we demonstrate a clear correspondence between isotope ratios of feathers and nonbreeding distributions of seven species from South Georgia tracked using loggers. This generated a rudimentary isoscape that was used to infer the habitat preferences of eight other species ranging in size from storm petrels to albatrosses, and which could be applied, with caveats, in other studies. Differences in inferred distribution within and between species had major implications for relative exposure to anthropogenic threats, including climate change and fisheries. Although there were no differences in isotope values between sexes in any of the smaller petrels, mean stable C (delta(13)C), but not stable N isotope ratios (delta(15)N), tended to be greater in females than males of the larger, and more sexually size-dimorphic species. This indicates a difference in C source (distribution), rather than trophic level, and a correspondence between the degree of sexual size dimorphism in Procellariiformes and the level of between-sex niche segregation.