Recent research on organizational mortality controls for the effect of age-varying organizational size and yields divergent results. Some studies find that ageing lowers mortality rates; others find the opposite pattern. We argue that this divergence reflects partly an overly simple specification of the effects of age and size. We argue that the effects of size on mortality rates differ by age group. Using complete data on organizational populations of automobile manufacturers in Britain, France, Germany, and the United States, we find that specifications with such age-variation improve over the usual specifications. The results for the American, French, and German populations indicate that age dependence is negative for the largest organizations and positive for small ones. The pattern is the reverse in the British population.