Data from the World Fertility Survey for selected Latin American countries are used to produce estimates of the simultaneous effects of breastfeeding and pace of childbearing on mortality during infancy and between first and fifth birthday. This is done by postulating models which take into account the reciprocal influences between the dynamics of birth intervals and breastfeeding. We also attempt to show that the effects vary according to several important characteristics of the child, mother, or community of residence. Finally, we investigate possible pitfalls in the inferences drawn by using alternative measurements of the main variables and by applying competing methods for the estimation of their effects. Although the results we obtain are quite robust to the definition of several indices and to the type of estimation method used, they remain partially inconclusive as a result of lack of proper controls for past and current health status of the infant.