OBJECTIVE: To test the hypothesis that use of thiazide diuretics prevents hip fracture and to study the risk for hip fracture associated with furosemide use. DESIGN: A case-control study. SETTING: Hospitals owned by a health maintenance organization in Washington. PATIENTS: Elderly patients (n = 462) hospitalized because of a hip fracture between 1977 and 1983 and an equal number of age- and sex-matched population-based control patients. MEASUREMENTS: Use of thiazide diuretics and furosemide was ascertained from medical records and computerized pharmacy records. The relative risk for hip fracture associated with diuretic use was calculated and adjusted for the potentially confounding effects of nursing home residence; previous hospitalizations; a history of stroke, alcoholism, or the organic brain syndrome; body weight; leg paralysis; and use of phenobarbital, corticosteroids, or other diuretics. Current and former users of diuretics were analyzed separately. MAIN RESULTS: The adjusted risk for hip fracture was 1.6 (95% CI, 1.0 to 2.5) for current thiazide users. The adjusted risk for hip fracture for current furosemide use was 3.9 (CI, 1.5 to 10.4). CONCLUSIONS: According to this study, use of thiazide diuretics did not protect against hip fracture and cannot be recommended for fracture prevention. Current furosemide use was also associated with hip fracture.