We examine nutritional outcomes in host communities exposed to refugee camps within a multi-country difference-in-differences framework across sub-Saharan Africa (SSA). Our study uses new spatially explicit data on refugee camps in SSA merged with the Demographic and Health Surveys to estimate the cross-country average treatment effect of camps. To test against bias in the coefficients of interest under staggered treatment timing, we use a diagnostic test to evaluate treatment effect homogeneity. We find that being within 10 kilometers of a camp decreases children’s weight-for-age z-scores by 10 percent of the sample mean. Children with married household heads experience improved nutrition outcomes near camps. We consider adult loss of employment and worsening child health as explanatory mechanisms: we find no significant evidence of worsening child health or a reduction in employment opportunities. We argue that rising child malnutrition among hosts is due to the changing composition of the host population or to price shocks under localized price dispersal.