Wild House mice, Mus musculus, were bred in two laboratory environments, one warm (controls) and one cold (Eskimo). At the seventh generation, mice of both stocks were cross-fostered at birth in both environments. In the warm environment, differences in both genotype and nest environment influenced growth: (1) Eskimo reared by Eskimo females were the heaviest of the four classes of fostered young; and (2) control foster parentage retarded growth. There was, however, no good evidence of differences in the reproductive performance of the four classes of fostered mice. In the cold environment, the effects of both genetical differences and of fostering were greater. Both the superior growth of Eskimo reared by Eskimo and the retarding effect of control foster parentage were more marked. Moreover, adult males with control foster parents had less fat than had those with Eskimo foster parents. Reproductive performance was also affected: (1) the young of the pairs with Eskimo genotype were heavier than the young of control pairs; (2) the litters of mice with Eskimo foster parents were larger than those of mice with control foster parents, and their young were heavier. Differences among the young of fostered mice represent a grandmother effect. Evidently, selection in a cold environment had led, not only to adaptive genetical changes in the ability to respond directly to cold, but also to changes in parental performance; and the latter enhanced the fitness, in the cold environment, of their offspring and grandoffspring.