Some data suggest that cesarean section reduces mother-to-child HIV-1 transmission. To assess the influence of mode of delivery and other maternal and infant factors on the rate of transmission, we analyzed the data of 1,624 children prospectively followed from birth. Of these, at the last visit 1,033 were >18 months of age or would have been had they not died of HIV-related illness. Among the 975 first singleton children, 180 18.5%; 95% confidence limits (CL), 16.1-20.9 acquired infection, as did 8 of 56 (14.3%; 95% CL, 5.1-23.5) second-born children. Multivariate stepwise analysis showed that vaginal delivery and development of symptoms in the mother were significantly and independently associated with a higher transmission rate (vaginal delivery: odds ratio, 1.69; 95% CL, 1.14-2.5; symptoms: odds ratio, 1.61; 95% CL, 1.12-2.3). In contrast, a history of maternal drug use, birth weight, breast-feeding (only 37 infants were breast-fed), and child's sex did not have a significant impact on viral transmission. The percentage of infected children was highest (30.7%) among very premature infants ( less than or equal to 32 weeks of gestation); this significant trend subsequently decreased to 11.9% at the week 42 (p < 0.001), suggesting a parallel reduction in peripartum transmission. The reduced rate of infection observed in infants born by cesarean section underlines the urgent need for randomized controlled trials to evaluate the protective role of surgical delivery in preventing perinatal HIV-1 transmission.