A survival analysis regression model is described for analyzing twin data on the age-at-menarche. The model includes latent genetic and environmental covariates and allows one to test hypotheses regarding the nature of familial aggregation for age-at-onset. Additionally, the model accommodates a variety of baseline survival distributions and therefore may be used to test different developmental hypotheses. Model-fitting results indicate that a survival model with a baseline gamma distribution gives an adequate fit to recalled age-at-menarche of 1,888 pairs of Australian female monozygotic and dizygotic twins. Further, results show that additive genetic and dominance genetic effects contribute to shared variation in age-at-menarche. If there are common environmental influences on the timing of menarche, they are completely obscured by nonadditivity in genetic factors, and information from other relationships would be required to detect their effect.