Background In 2019, 1 million in utero-HIV-exposed but -uninfected children (CHEU) were born in sub-Saharan Africa. In the absence of breastfeeding and maternal antiretroviral therapy (ART), HIV-exposed uninfected children (CHEU) have higher risks of mortality, growth faltering, infectious morbidity, and developmental delay than the general population, but data are limited on these outcomes among breastfed CHEU born to women who received universal (not restricted by disease severity) ART in pregnancy. This thesis addresses these knowledge gaps by comparing health outcomes of breastfed CHEU to breastfed CHU in the first year of life. Methods This research incorporates data from two prospective, linked cohort studies of pregnant women living with HIV (Maternal Child Health Antiretroviral, MCH-ART study, 2013-2016) and without HIV (HIV-unexposed uninfected, HU2 study; 2014-2017) in Gugulethu, Cape Town, South Africa. Enrolled at first antenatal visit at a primary care clinic, women were followed-up during pregnancy and postpartum with their breastfed infants, through 12 months with ~3-monthly study visits. Study staff administered questionnaires addressing maternal and child health, infant feeding, psycho-social and behavioural factors; and measured maternal-infant anthropometry [expressed as Z-scores for age: weight-for-age, WAZ; length-for-age, LAZ; head circumference-for-age, HCAZ]; WLHIV provided blood for repeated HIV viral load. At 12 months, the Bayley Scales of Infant Development, 3rd edition (BSID-III) was used to assess neurodevelopment. Findings WLHIV (100% antenatal ART) reported worse living conditions and higher risks of alcohol use and intimate partner violence than HIV-negative women. Similar proportions of CHEU and CHU were born preterm (11%) or small-for-gestational-age (10%). Exclusive breastfeeding was more common among CHEU than CHU, but overall duration of breastfeeding was shorter among CHEU. However, unless otherwise reported, adjustment for confounders did not change inferences below. In analysis of child growth, weight and head circumference trajectories were similar for CHEU and CHU from 6 weeks to 12 months. Both groups exhibited rapid weight gain with increasing WAZ over time; by 12 months, almost one-fifth of all children were overweight. Length trajectories for CHEU and CHU diverged after 6 months, with onset of linear growth faltering occurring earlier and more rapidly among the CHEU; by 12 months, stunting risk was doubled among CHEU vs CHU. Stratified by birth size, differences in LAZ between CHEU and CHU were magnified for those born small-forgestational age and absent for those born appropriate-for-gestational age. Infectious morbidity analyses revealed greater risks among CHEU than CHU in the first 6 months with not thereafter. Between 7 days and 3 months of life, CHEU (vs CHU) experienced three times more infection-related hospitalisations; rates for CHEU with healthier mothers (lower viral load, higher CD4 count, ART started early in pregnancy) approximated those of CHU, while CHEU of mothers with late ART initiation and advanced disease had four-fold more infectious-cause hospitalization. Breastfeeding and complete vaccinations were protective. At 12 months, mean composite cognitive, motor and language scores were within normal range and similar for both groups. Overall, risks of any developmental delays were low but slightly higher among CHEU than CHU in cognitive and motor domains. Compared to term HU, term HEU children had similar odds of motor delay, preterm HU children had 5-fold increased odds of delay and preterm HEU children, 16-fold. In CHEU, cumulative maternal viremia was associated with lower average scores and increased risk of moderate delays in motor and language domains. Conclusion Subtle health outcome differences persisted between CHEU and CHU despite breastfeeding and universal maternal ART in pregnancy. Reassuringly, the magnitudes of differences were small and predominantly associated with preventable factors including late ART initiation, advanced maternal disease stage, lack of breastfeeding, and incomplete vaccination. CHEU born too soon or too small were at highest risk of adverse outcomes, suggesting fetal origins of disease in the context of maternal HIV.