Abstract Inbreeding depression was simultaneously studied under contrasted environments, laboratory and natural conditions, using individuals originating from 14 families of the freshwater snail Physa acuta. Both survival and growth of juveniles showed inbreeding depression under laboratory conditions. The same fitness components were monitored with mature snails either kept under laboratory conditions or released at a natural site and analysed using capture-mark-recapture models. Genetic composition of both samples was similar. Inbreeding depression on survival was highest in the laboratory while strong outbreeding depression was revealed in the field. Thus inbreeding depression may not be always higher under natural conditions, at the opposite of what is commonly assumed. We suggest that inbreeding depression is dependent on metabolic requirements imposed by the environment. Other evidences showing that inbreeding depression is environment-dependent are reviewed. We conclude that genetic models should include both genetic and environmental variance in inbreeding depression for studying mating system evolution.