Data from the history of Havasupai fertility and fertility data on other populations indicate that variables other than stochastic flux and differential mortality may affect sex ratios in human populations. They suggest that sex ratios at birth may vary with birth order in a conjugal union, maternal age, form of marriage, and political relations between the sexes. The behavioral variable which appears to tie all these factors to sex ratio at birth is coital frequency. The bases of this connection remain obscure, but its consequences are evident and important. Not only do social practices through their impact on coital rates lead to variation, but that variation produces more variation. When sex ratio at birth is correlated with birth order and maternal age, a population's sex ratio at birth will reflect its age/sex-specific fertility and mortality rates. Further, because the process which ties the population sex ratio at birth to those rates is recursive, the process may generate systemic historic oscillations in the population's age/sex structure even when vital rates are stable.