OBJECTIVES: First, to compare the prevalence of HIV infection among women in the general population and antenatal clinic (ANC) attenders in rural Mwanza, Tanzania, and second, to validate a method for adjusting HIV prevalence in ANC attenders to estimate the prevalence in the general female population aged 15-44 years. METHODS: A cross-sectional population survey was conducted in 12 rural communities of Mwanza Region between 1991 and 1992. From the same communities sequential ANC attenders were recruited on two occasions between 1991 and 1993. Consenting subjects were interviewed, examined, treated and a serum sample was tested for HIV. The HIV prevalence in women in the general population was compared with unadjusted and adjusted prevalences in ANC attenders. Parity-adjusted prevalences were obtained by applying correction factors to the observed prevalences in parous and nulliparous ANC attenders. RESULTS: A total of 5675 women aged 15-44 years from the general population and 2265 ANC attenders had complete socio-demographic and laboratory data. Unadjusted HIV prevalence was significantly lower in ANC attenders (3.6%) than women from the general population (4.7%, P = 0.025), but after adjustment there was no significant difference between the two groups (4.6 versus 4.7%, P = 0.95). CONCLUSION: In this rural population, the HIV prevalence in ANC attenders underestimated the prevalence among women in the general population, but this difference was eliminated by applying parity-based correction factors. Information on parity should be routinely collected in ANC-based HIV sentinel surveillance.