Objectives:To evaluate the impact of perinatal zidovudine use on the risk of perinatal transmission of HIV and to determine risk factors for transmission among women using perinatal zidovudine.Design:Prospective cohort study of 1533 children born to HIV-infected women between 1985 and 1995 in four US cities.Methods:The association of potential risk factors with perinatal HIV transmission was assessed with univariate and multivariate statistics.Results:The overall transmission risk was 18% [95% confidence interval (CI), 16–21]. Factors associated with transmission included membrane rupture > 4 h before delivery [relative risk (RR), 2.1; 95% CI, 1.6–2.7], gestational age < 37 weeks (RR, 1.8; 95% CI, 1.4–2.2), maternal CD4+ lymphocyte count < 500 × 106cells/l (RR, 1.7; 95% CI, 1.3–2.2), birthweight < 2500 g (RR, 1.7; 95% CI, 1.3–2.1), and antenatal and neonatal zidovudine use (RR, 0.6; 95% CI, 0.4–0.9). For infants exposed to zidovudine antenatally and neonatally, the transmission risk was 13% overall but was significantly lower following shorter duration of membrane rupture (7%) and term delivery (9%). The transmission risk declined from 22% before 1992 to 11% in 1995 (P< 0.001) in association with increasing zidovudine use and changes in other risk factors.Conclusions:Perinatal HIV transmission risk has declined with increasing perinatal zidovudine use and changes in other factors. Further reduction in transmission for women taking zidovudine may be possible by reducing the incidence of other potentially modifiable risk factors, such as long duration of membrane rupture and prematurity.