A prospective study on 289 women with clinical stage I cutaneous melanoma was done to determine the relationship between estrogen administration, tumor characteristics, and survival. Eighty-two women took oral contraceptives (OC) and 44 took menopausal estrogens (MPE) prior to the diagnosis of melanoma. Users of OC presented with thinner primary tumors than nonusers of OC (P less than .01). A similar trend was observed in users of MPE. Women who used OC in the year prior to the diagnosis of melanoma had statistically thinner tumors than those who had discontinued use of OC more than 1 year prior to diagnosis (P less than .025). A statistically significant preponderance of truncal lesions was observed among users of OC (P less than .01). Other tumor characteristics were unaltered by estrogen administration. Duration of use and time in relation to diagnosis of melanoma did not affect survival. Women who took hormones had slightly better 5- and 9-year survival rates than nonusers. These results suggest that prior estrogen use and, particularly, use of OC in women developing melanoma have no deleterious effect.