OBJECTIVE: To determine whether data from voluntary counseling and testing (VCT)/prevention of mother-to-child transmission (PMTCT) programs can be used for HIV surveillance. METHODS: Women attending an antenatal clinic at the district hospital in Entebbe, Uganda, from May 2002 to April 2003 were offered counseling and HIV testing with same-day results (VCT) and nevirapine for PMTCT was provided for HIV-positive women and their babies. Those who declined VCT were tested for HIV anonymously. RESULTS: Overall, 2635 women accepted VCT; 883 were tested anonymously. HIV prevalence was higher in VCT than in anonymously tested women in the first month of the program (20% vs. 11%, P=0.05) and in months with <70% VCT uptake (17% vs. 8%, P<0.001) but was similar in months with high uptake. Uptake of VCT was higher in women who had risk factors for HIV, especially those who believed themselves to have been exposed (84% vs. 73%, P<0.001). CONCLUSION: There was a bias to accepting VCT in women with HIV, or risk factors for HIV infection, the former most apparent when there was low coverage. Data from VCT/PMTCT programs cannot replace anonymous surveillance for monitoring of HIV epidemic trends where coverage is incomplete within clinics or communities.