Estrogen receptor (ER) and progestin receptor were measured in samples of tumors obtained at first laparotomy from 97 previously untreated patients suffering with a primary ovarian epithelial tumor, for whom a 3-year follow-up was available. The presence or absence of steroid receptors (threshold arbitrarily fixed at 10 fmol/mg of cytoplasmic protein) was determined by the dextran coated charcoal method and related to a number of patient characteristics such as the residual disease (cutoff, 2 cm), histological type, International Federation of Gynecologists and Obstetricians grade and stage, and age. Results were analyzed by univariate and multivariate methods. (a) The tumor ER positivity was associated with better survival; progestin receptor showed a similar trend but did not reach statistical significance. (b) After stratification for residual tumor the association ER positivity/better survival was still statistically significant in the subset of patients with residual tumor greater than 2 cm. (c) When the median survival times were considered it became apparent that progestin receptor absence nullified the effect associated with positive ER. (d) Multivariate analysis confirmed that among the variables considered the main determinants of prognosis were the size of the residual tumor, serous histological type, and positive ER.