The objective of this study was to examine the association of HIV-1 infection with rates of pregnancy and pregnancy loss in Dar es Salaam, Tanzania. A retrospective cohort study of 1,006 HIV-infected women and 485 uninfected women was employed. In multivariate analyses controlling for other predictors of pregnancy, the association of HIV-seropositivity with a woman's reported number of pregnancies was of borderline significance (RR = 1.13, 95% CI = 1.00, 1.27). HIV infection was not associated with pregnancy loss in multivariate analysis. The adjusted pregnancy rate ratio comparing HIV-positive women at the earliest stages of infection to all uninfected women was 1.22 (95% CI = 1.04, 1.42). HIV infection was not associated with female fertility when comparing women in the most advanced stages of infection to all uninfected women. We conclude that HIV-1 infected women had higher pregnancy rates than uninfected women. This association disappeared when analyses were limited to women with advanced disease. Abstract word count: 150.