We conducted independent case-control and retrospective cohort investigations to assess the relationship between diaphragm use and urinary tract infection (UTI). In the former, we compared diaphragm use and vaginal flora among 114 women with acute UTI and 85 women with acute urinary tract symptoms and no UTI. In the latter study, we ascertained the incidence of UTI in 192 diaphragm users and 182 women taking oral contraceptives during a mean follow-up of 9.4 months. Both studies demonstrated a significantly increased risk of UTI in diaphragm users: relative odds were 2.0 in the case-control study and the relative risk was 2.5 in the retrospective cohort study. Vaginal colonization with Escherichia coli was significantly greater in diaphragm users. The incidence of UTI in the cohort study was 26.6 per 1,000 patient-months for diaphragm users and 8.9 per 1,000 patient-months for women taking oral contraceptives. The increased risk of UTI in diaphragm users could not be attributed to differences in age, parity, sexual activity, or previous UTI.