Seven trials were included in a Cochrane review . Four trials comparing zidovudine with placebo involving 1585 participants were included. Compared with placebo, there was a significant reduction in the risk of mother-to-child transmission with any zidovudine (RR 0.54, 95% CI 0.42 to 0.69). There is no evidence that 'long course therapy' is superior to 'short course therapy'. One trial compared intrapartum and postnatal nevirapine with intrapartum and postnatal zidovudine in 626 women, the majority of whom breast fed their infants. Compared with zidovudine, there was a significant reduction in the risk of mother-to-child transmission of HIV with nevirapine (RR 0.58, 95% CI 0.40 to 0.83). One trial comparing elective caesarean section with anticipation of vaginal delivery involving 436 participants was included. Compared with vaginal delivery, there was a significant reduction in the risk of mother-to-child transmission of HIV infection with caesarean section (RR 0.17, 95% CI 0.05 to 0.55). One trial comparing hyperimmune immunoglobulin plus zidovudine with non-specific immunoglobulin plus zidovudine involving 501 participants was included. The addition of hyperimmune immunoglobulin to zidovudine does not appear to have any additional effect on the risk of mother-to-child transmission (RR 0.67, 95% CI 0.29-1.55). .