To assess the effect of postmenopausal use of estrogens on the subsequent risk of hip fracture, we performed a retrospective cohort study of 2873 women in the Framingham Heart Study. Information obtained at routine biennial examinations about the use of estrogens, body weight, age at menopause, smoking, and alcohol consumption was used to evaluate the risk of hip fracture among postmenopausal women who received estrogens. Hip fractures occurred in 179 postmenopausal women, at a rate that increased exponentially after the age of 50. The risk of fracture was inversely related to weight at all ages. The relative risk of hip fracture in subjects who had taken estrogens at any time was 0.65 after adjustment for age and weight (95 percent confidence interval, 0.44 to 0.98). The adjusted relative risk in women who had taken estrogens within the previous two years was further reduced, to 0.34 (95 percent confidence interval, 0.12 to 0.98). Taking estrogens within four years of menopause also protected against fracture. The number of women in each age group who took estrogens was insufficient for a definitive evaluation of risk, but recent use of estrogens appeared to be protective in women under the age of 65 (no fractures among those who took estrogens) and those 65 to 74. We cannot exclude some degree of selection bias among the women who received estrogen-replacement therapy. Nevertheless, this large cohort study supports the hypothesis that postmenopausal use of estrogens protects against subsequent hip fracture in women.