OBJECTIVE: To assess morbidity and mortality in HIV-exposed uninfected (HEU) children to help guiding appropriate clinical care and effective preventive interventions. DESIGN: This is a longitudinal study comparing two cohorts of children; one born to HIV-infected women and the other born to HIV-uninfected women. METHODS: We have analyzed prospectively obtained information on nutritional status, morbidity and mortality from 966 HEU and 909 HIV-unexposed infants followed up until their first 18 months of life at a referral health facility in southern Mozambique. Determinants for adverse health outcomes in HEU children were also assessed using multivariate logistic regression. RESULTS: Increased incidence of hospital admissions (P = 0.0015), shorter survival in the first 18 months of life (P = 0.0510) and moderate and severe malnutrition (P = 0.0006 and 0.0014, respectively) were observed among HEU children compared with HIV-unexposed children. Incidence of outpatient attendance in HEU children was associated with being men, older age and the mother being on antiretroviral treatment. Among HEU children, those who were never breastfed, or who were weaned or were partially breastfed, had an increased incidence of hospital admissions compared with children who were exclusively breastfed. CONCLUSION: Maternal HIV infection has important health consequences in non-HIV-infected children. As the prevalence of HIV-infected pregnant women is maintained and the proportion of HIV-infected children declines because of the scale-up of antiretroviral treatment during pregnancy and breastfeeding, more focus should be given to the health needs of HEU children to ensure that the post-2015 sustainable development goals are met.