Objective: To describe vaginal bleeding and associated pain in pregnancies terminated by childbirth. Design: A descriptive study based on information obtained by questionnaires administered during pregnancy. Setting: Antenatal care clinic at the Obstetrical Department, Aarhus University Hospital, Denmark. Subjects: Women (n = 8714) with singleton pregnancies attending routine antenatal care at the Department during a 2-year period from 1989 to 1991. Three questionnaires were administered during pregnancy and 5868 (67%) women completed all questionnaires. Main measures: Bleeding was characterized by frequency, gestational age at first occurrence, duration, and amount. Hemorrhage associated with abdominal pain and hospitalization was also described. Results: The overall frequency of bleeding was 19% (n = 1091). The median week of first occurrence was 8 gestational weeks. The median duration of bleeding was 2 days, and the median number of episodes was one. An association between the amount and duration of bleeding was found. Two thirds of the women had no abdominal pain associated with bleeding. Women with pain experienced heavier bleeding and bled for more days than women without pain. Bleeding resulted in hospitalization in one out of five women. In particular, women with second trimester bleeding, with heavy bleeding, long duration, and associated pain were hospitalized. Conclusion: By the use of self-administered questionnaires a higher frequency of bleeding was found than in previous studies in which information has been based on medical records. Thus, the source of information in studies of pregnancy and bleeding-associated pregnancy complications must be carefully considered and described.