BACKGROUND: HPTN 046 compared the efficacy and safety of infant nevirapine (NVP) among HIV-exposed breastfed infants randomized at 6 weeks to 6 months to t NVP or placebo to prevent postnatal infection: we report final 18-month outcomes. METHODS: Randomized, placebo-controlled trial in 4 African countries. Infant diagnostic HIV testing was performed regularly from birth through 18 months. Kaplan-Meier analysis was used to assess 18-month cumulative infant HIV infection, HIV infection/or death, and mortality rates. RESULTS: Between 6 weeks and 6 months, postnatal HIV infection rates were significantly lower among infants receiving daily NVP from 6 weeks to 6 months 1.1% [95% confidence interval (CI): 0.2% to 1.8%], compared with placebo 2.4% (95% CI: 1.3% to 2.6%), P = 0.049, but not significantly lower thereafter. Eighteen-month postnatal infection rates were low: 2.2% (95% CI: 1.1% to 3.3%) versus 3.1% (95% CI: 1.9% to 4.4%), respectively, P = 0.28. Mortality and HIV infection/death did not differ between arms at any age. Infants of women receiving antiretroviral therapy (ART) for their own health had the lowest 18-month postnatal infection rates (0.5%, 95% CI: 0.0% to 1.1%). However, HIV infection/death rates at 18 months were not significantly different for infants of mothers on ART (3.7%, 95% CI: 1.9% to 5.5%), and infants of mothers with CD4 counts of ≥350 cells per cubic millimeter not receiving ART (4.8%, 95% CI: 2.7% to 6.8%; P = 0.46). There were no differences in adverse events between study arms. CONCLUSIONS: This trial demonstrated early but not late differences in postnatal HIV transmission among infants randomized at age 6 weeks to extended NVP or placebo, underscoring the importance of continued prophylaxis throughout breastfeeding.