Objectives: The extent to which HIV affects pregnancy-related mortality in countries with high HIV/AIDS and maternal mortality is poorly understood. The objectives of this study were to investigate the mortality of women of reproductive age by both HIV and pregnancy status, and quantify the excess mortality attributable to HIV during pregnancy in Pointe Noire, Congo. Design: Prospective mortuary investigation of all deaths in women aged 15-44 years, during 112 consecutive days. Methods: Mortality rates by HIV and pregnancy were computed. During the study period, 378 corpses were examined, blood was tested for HIV and pregnancy, relatives were interviewed and hospital files were reviewed. Denominators were obtained from a census with women-years assigned to pregnancy and/or HIV based on levels of fertility and HIV prevalence in the city. Results: The mortality rate was 32 times higher [95% confidence interval (CI), 25-39] among HIV-positive than among HIV-negative women. The relative increase in mortality associated with HIV was much higher in non-pregnant [rate ratio (RR), 41; 95% CI, 32-52] than in pregnant women (RR, 4; 95% CI, 2-9). Among HIV-positive women, pregnancy appeared to confer a survival benefit. Conclusion: These findings have important implications for the interpretation of trends in maternal mortality in the context of HIV. The apparent survival benefit of pregnant HIV-positive women is largely due to their low fertility in the latest stage of the disease. As the HIV epidemic matures and more women become severely ill, any potential adverse effects associated with HIV and pregnancy may be increasingly offset by selection effects, and maternal mortality may not increase further. © 2005 Lippincott Williams & Wilkins.