This paper estimates a model of dynamic intrahousehold investment behavior which incorporates family fixed effects and child endowment heterogeneity. This framework is applied to large American and British survey data on birth outcomes, with focus on the effects of antenatal parental smoking and maternal labor supply net of other maternal behavior and child characteristics. We find that maternal smoking during pregnancy reduces birth weight and fetal growth, while paternal smoking has virtually no effect. Mothers' work interruptions of up to two months before birth have a positive effect on birth outcomes, especially among British children. Parental behavior appears to respond to permanent family-specific unobservables and to child idiosyncratic endowments in a way that suggests that parents have equal concerns, rather than efficiency motives, in allocating their prenatal inputs across children. Evidence of equal concerns emerges also from the analysis of breastfeeding decisions, although the effects in this case are weaker.