Using household survey data for Guatemala, this paper investigates the role of water and sanitation infrastructure on diarrhea incidence in children. Hierarchical logit models of diarrhea incidence are estimated to account for potential regional heterogeneity of water and sanitation effects. Results indicate that the incidence probability of diarrhea is on average 20% lower in homes connected to a sewerage system. The effect of in-home access to tap water is weaker at 11% and subject to regional heterogeneity. Findings also indicate that consumption of bottled water reduces the incidence probability of diarrhea by 20–22%. In-home water treatments have no effect on incidence of diarrhea. Policy implications are discussed.