Objective. To demonstrate the feasibility and acceptability of introducing postpartum voluntary counselling and testing (PPVCT) and the provision of post-exposure nevirapine prophylaxis (PEP) to HIV-exposed infants whose mothers did not receive any antiretroviral prophylaxis to prevent mother-to-child transmission (PMTCT) in a large, tertiary hospital in the South African setting. Design. Observational, interventional study. Setting. The programme was implemented at Chris Hani Baragwanath Hospital (tertiary referral centre) in Soweto, South Africa, following a study that established the efficacy of a postpartum regimen of PEP in PMTCT. Participants. From January 2003 to December 2004, 7 500 women who delivered at Chris Hani Baragwanath Hospital without a documented HIV-1 result were identified in the postnatal wards. PPVCT was offered to all eligible participants. Intervention. On-site HIV rapid tests were performed on all women who agreed to testing. For those women testing HIV-1 positive, a single dose of nevirapine syrup was offered to their infants as PEP within 72 hours after delivery. Infant feeding counselling, assistance with follow-up care and support programmes were also offered. Main outcome. From January 2003 to December 2004, 34 776 deliveries occurred at Chris Hani Baragwanath Hospital. Of these, 7 500 (21.5%) had no documented HIV status. After delivery 5 751 (76.7%) women were offered VCT, and of these 3 794 (66%) accepted testing. Of these women, 1 294 (34%) tested HIV positive and 1 243 (96%) women accepted the administration of single-dose nevirapine to their infants. Conclusions. The uptake of PPVCT is comparable to that seen in established antenatal VCT despite the numerous challenges PPVCT presents. This suggests that PPVCT is both an acceptable and a feasible option in a busy, resource-limited setting and remains an important strategy in PMTCT in untreated individuals.