The social insects have always posed an evolutionary puzzle to biologists. They have evolved worker castes which rarely or never reproduce. But how can genes that prevent reproduction ever become common in populations? The answer is that workers aid the reproduction of relatives who, though they do not express the genes for worker behavior, do possess them and transmit them to their offspring. A combination of theory and empirical evidence suggests that the aid given must be considerable: workers must generally increase the reproduction of their relatives by more than the amount that they decrease their own reproduction. How such reproductive efficiencies of scale can be achieved is the main theme of the proposed research, focusing on primitively social insects that have not evolved caste specializations. Mathematical models will show how simple and plausible patterns of mortality and reproductive timing can lead to these efficiencies of scale. Field studies of wasps will serve to test the models. The result of this work will be a better understanding of why some species engage in altruistic behavior and why others do not. These insights may lead to the recognition of similarities underlying the evolution of altruism in the social insects and higher animals.