This paper examines the two-way relationship between birth interval and child survival and compares the behaviour of households in the Indian and Pakistani provinces of Punjab. Birth interval and child survival are modelled here as correlated hazard processes, allowing for mother- specific unobserved heterogeneity. We find evidence of significant mutual dependence between birth interval and child survival in both samples. There are also significant differences between Indian and Pakistani households. Part of the difference in behaviour could be explained by differences in female literacy, which in turn highlight the differences in religion and state policies in these two neighbouring states.