Objectives: To compare rates of reproductive events before and after HIV diagnosis in a cohort of women with HIV infection, and to consider the impact of HIV diagnosis on the outcome of pregnancy. Design: Observational cohort study of 503 women recruited from 15 genitourinary medicine/HIV clinics in Britain and Ireland. The 503 women had 580 pregnancies before diagnosis of HIV infection and 202 after HIV diagnosis. Methods: Using date of birth, date of HIV diagnosis, the outcome of all lifetime pregnancies and date of each outcome, age-specific rates (per 100 women-years) of pregnancy, miscarriage, termination and live-birth were calculated before HIV diagnosis, and separately after HIV diagnosis. Rates after HIV diagnosis were age-standardized for comparison with rates before HIV diagnosis. Rates were also calculated separately by ethnic group and HIV transmission group. Results: In women aged 20-34 years, the age-adjusted live-birth rate fell by 44% from 10.2 [95% confidence interval (CI), 9.2-11.2] per 100 women-years before HIV diagnosis to 5.7 (95% CI, 4.3-7.1) after diagnosis. Most of the decline reflected an increase in termination rate from 3.5 (95% CI, 2.9-4.1) before HIV diagnosis to 6.3 (95% CI, 4.7-7.9) after diagnosis. A decline in live-births together with a rise in termination after HIV diagnosis was a consistent finding across age and ethnic groups. However, black African women had the smallest reduction in live-births, despite the greatest increase in termination, because the pregnancy rate increased after HIV diagnosis in this group. Conclusions: Diagnosis of HIV infection in women has a substantial impact in reducing live-birth rates. These findings have important implications for expanding HIV testing in women. They highlight the need for better understanding of reproductive decision-making in the context of HIV infection and better contraceptive support for HIV-infected women and their partners.