Female residents of six counties in Washington and Utah in whom ovarian cancer was diagnosed during 1975-79 were interviewed concerning prior use of noncontraceptive estrogens. Interviews with a random sample of women drawn from these counties were obtained for comparison. Taken as a whole, women with epithelial ovarian tumors reported estrogen use for 1 year or more somewhat more often than did controls [relative risk (RR) = 1.3, 95% confidence interval (CL) = 0.9-1.8]. The proportion of women who had used estrogens varied according to the histology of their tumor; notably, of 17 women with endometrioid tumors, 12 had used estrogens (RR = 3.1, 95% CL = 1.0-9.8). The excess risk of endometrioid carcinoma related to estrogen use was not further increased when these hormones were taken for long durations, in high doses, or proximate to the time of diagnosis. Neither was there any evidence of increased risk from estrogen use in the small group of women with clear cell tumors, a histologic tumor type believed to be closely related to endometrioid carcinoma. Thus the observed association between menopausal estrogen use and the occurrence of ovarian endometrioid tumors should be interpreted cautiously until the results of additional studies are available.