BACKGROUND: Single-dose nevirapine (SDNVP) for the prevention of mother-to-child HIV transmission (PMTCT) results in the selection of resistance mutants among HIV-infected mothers. The effects of these mutations on the efficacy of SDNVP use in a subsequent pregnancy are not well understood. METHODS: We compared risks of perinatal HIV transmission between multiparous women who had previously received a dose of SDNVP (exposed) and those that had not (unexposed) and who were given SDNVP for the index pregnancy within a PMTCT clinical study. We also compared transmission risks among exposed and unexposed women who had two consecutive pregnancies within the trial. Logistic regression modeling was used to adjust for possible confounders. RESULTS: Transmission risks did not differ between 59 SDNVP-exposed and 782 unexposed women in unadjusted analysis or after adjustment for viral load and disease stage (adjusted odds ratio 0.6, 95% confidence interval [CI] 0.2 to 2.0). Among 43 women who had two consecutive pregnancies during the study, transmission risks were 7% (95% CI 1% to 19%) at both the first (unexposed) and second (exposed) delivery. The results were unchanged, if infant death was included as an outcome. CONCLUSION: These data suggest that the efficacy of SDNVP may not be diminished when reused in subsequent pregnancies.