Lifelong antiretroviral therapy (ART) provision to all pregnant HIV-positive women (“Option B+”) has been recommended by the World Health Organization since 2013, but there remain limited data on the effects of Option B+ on long-term HIV-free survival in breastfeeding HIV-exposed infants. The Kigali Antiretroviral and Breastfeeding Assessment for the Elimination of HIV (Kabeho) study enrolled HIV-positive women from the third trimester of pregnancy to 2 weeks postpartum in 14 heath facilities implementing Option B+ in Kigali, Rwanda. Mother–child pairs in the longitudinal observational cohort were followed until 24 months postpartum, with HIV diagnostic testing at 6 weeks, and 9, 18 and 24 months. The Kaplan–Meier method was used to estimate HIV transmission, survival, and HIV-free survival through 24 months. We enrolled 608 HIV-positive women in 2013–2014; birth outcome data were available for 600 women and 597 live-born infants. By 6 weeks, 11 infants had died and 3 infants had confirmed HIV infection (0.5% transmission; 95% confidence interval [CI] 0.2–1.6). At 9 months, there were 9 additional deaths and 2 new infections (cumulative transmission 0.9%, 95% CI 0.4–2.2). At 18 months, there were 6 additional deaths and no new infant infections. At 24 months, there were no additional child deaths and 1 new infection (cumulative 2.2%, 95% CI 0.7–7.0), for an overall 24-month HIV-free survival of 93.2% (95% CI 89.5–95.6). Low transmission rates and high HIV-free survival at 24 months were achieved in breastfeeding infants of HIV-positive mothers receiving universal ART in urban health facilities in Rwanda, though vigilance on maintaining viral suppression for ART-experienced women is needed.