BACKGROUND: Intrapartum single-dose nevirapine plus third trimester maternal and infant zidovudine are essential components of programs to prevent mother-to-child transmission of human immunodeficiency virus (HIV) in resource-limited settings. The persistence of nevirapine in the plasma for 3 weeks postpartum risks selection of resistance mutations to nonnucleoside reverse-transcriptase inhibitors (NNRTIs). We hypothesized that a 1-month zidovudine-didanosine course initiated at the same time as single-dose nevirapine (sdNVP) would prevent the selection of nevirapine-resistance mutations. METHODS: HIV-infected pregnant women in the PHPT-4 cohort with CD4 cell counts >250 cells/mm3 received antepartum zidovudine from the third trimester until delivery, sdNVP during labor, and a 1-month zidovudine-didanosine course after delivery. These women were matched on the basis of baseline HIV load, CD4 cell count, and duration of antepartum zidovudine to women who received sdNVP in the PHPT-2 trial (control subjects). Consensus sequencing and the more sensitive oligonucleotide ligation assay were performed on samples obtained on postpartum days 7-10, 37-45, and 120 (if the HIV load was >500 copies/mL) to detect K103N/Y181C/G190A mutations. RESULTS: The 222 PHPT-4 subjects did not differ from matched control subjects in baseline characteristics except for age. The combined group median CD4 cell count was 421 cells/mm3 (interquartile range [IQR], 322-549 cells/mm3), the median HIV load was 3.45 log10 copies/mL (IQR, 2.79-4.00 log10 copies/mL), and the median duration of zidovudine prophylaxis was 10.4 weeks (IQR, 9.1-11.4 weeks). Using consensus sequencing, major NNRTI resistance mutations were detected after delivery in 0% of PHPT-4 subjects and 10.4% of PHPT-2 controls. The oligonucleotide ligation assay detected resistance in 1.8% of PHPT-4 subjects and 18.9% of controls. Major NNRTI resistance mutations were detected by either method in 1.8% of PHPT-4 subjects and 20.7% of controls (P