Antiretroviral drugs (ARVs) have been shown to be efficacious in decreasing mother-to-child transmission (MTCT) of HIV. A summary estimate of the efficacy of ARVs in reducing MTCT is important for modeling and policy decisions. However, no one has hitherto attempted to generate this summary estimate for Africa, the continent with the greatest HIV/AIDS burden. This study estimates the efficacy of ARVs in reducing MTCT in Africa through a meta-analysis of published studies conducted in Africa. Using an a priori protocol, Medline, EMBASE, and the Cochrane Library were searched for primary studies that measured MTCT of HIV, had ARVs as the exposure to the mother, and were conducted in Africa. Extracted data included characteristics of the study, population, quality, exposure, and results. The data were analyzed using a random effects model with each trial arm as a data point. Ten randomized clinical trials conducted in West, East, and Southern Africa published from 1999 to 2007 satisfied the inclusion criteria. They ranged in sample size from 139 to 1797, and used different ARV regimens as the exposure to the mother antepartum, intrapartum, or postpartum, and to the baby. The combined effect estimate of using ARVs is 10.6% (95% CI: 8.6-13.1) transmission at 4-6 weeks and 21.0% (95% CI: 15.5-27.7) transmission for placebo. This represents approximately 50% efficacy. The result is stable and not driven by any single study. All regimens were well tolerated. We conclude that ARV use to reduce MTCT of HIV in Africa is efficacious and well tolerated.