Much evidence of HIV-exposed uninfected (HEU) infant infectious morbidity predates availability of maternal combination antiretroviral therapy and does not control for universal risk factors (preterm birth, low birth weight, suboptimal breastfeeding and poverty). This prospective cohort study identified HIV-infected and HIV-uninfected mothers and their newborns from South African community midwife unit. The primary outcome, infectious cause hospitalization or death before 6 months of age, was compared between HEU and HIV-unexposed (HU) infants and classified for type and severity using validated study-specific case definitions. Adjusted odds ratios (aORs) were calculated by logistic regression including stratified analyses conditioned on breastfeeding. One hundred and seventy-six (94 HEU and 82 HU) mother-infant pairs were analyzed. HIV-infected mothers were older (median, 27.8 vs. 24.7 years; P < 0.01) and HU infants more often breastfed (81/82 vs. 35/94; P < 0.001). Groups were similar for maternal education, antenatal course, household characteristics, birth weight, gestational age and immunizations. The primary outcome occurred in 17 (18%) HEU and 10 (12%) HU infants [aOR, 1.45; 95% confidence interval (CI): 0.44-4.55]. In stratified analysis restricted to breastfed infants, the aOR for hospitalization due to very severe infection or death was 4.2 (95% CI: 1.00-19.2; P = 0.05) for HEU infants. Hospitalization for diarrhea was more common in HEU than HU infants [8/94 (8.5%) vs. 1/82 (1.2%); P = 0.04]. The difference between HEU and HU infants in the probability of infectious cause hospitalization or death in the first 6 months of life was not significant. However, among breastfed infants, severe infectious morbidity occurred more often in HEU than HU infants.