OBJECTIVE: To assess the impact of HIV and antiretroviral exposure without infection on lung growth and function over the first 2 years of life. DESIGN: Prospective observational study of an African birth cohort, Drakenstein Child Health Study. METHOD: Infants enrolled antenatally had lung function measured at 6 weeks, 1 and 2 years. HIV-infected women received antiretroviral therapy (ART) as per local guidelines. The association between HIV and antiretroviral exposure with lung function was assessed using mixed effects modelling. RESULTS: Of 1143 infants born, two HIV-infected infants were excluded from analysis; 909 (80%) infants had lung function collected at 6 weeks [190 (21%) were HIV-exposed uninfected (HEU)]; 782 (69%) at 1 year and 741 (65%) at 2 years. At 6 weeks HEU infants had larger tidal volume compared with HIV-unexposed infants (1.13 ml, confidence interval: 0.02-2.23, P = 0.045). High maternal viral load was associated with a 17% lower expiratory flow over 2 years (0.17, confidence interval 0.00-0.34, P = 0.046). First-line ART initiated during pregnancy was associated with lower infant tidal volume at 6 weeks compared with those who initiated ART before pregnancy (-2.7 ml, -5.31 to -0.10, P = 0.042), and low maternal CD4 cell counts associated with lower infant tidal over 2 years (-11.1 ml, -18.58-3.58, P = 0.004). CONCLUSION: HIV exposure is associated with altered lung function in early life, with a vulnerable HEU subgroup based on maternal disease severity, immunological compromise and ART exposure. These data highlight the importance of ongoing surveillance of respiratory health in HEU children.