Positive results are reported from a double-blind study of estrogen therapy administered to severely depressed, inpatient women who had failed to respond to various conventional treatments of depression. Large doses of oral conjugated estrogen were administered for a three-month period to 23 premenopausal and postmenopausal inpatient women. Placebos were administered for a comparable period to 17 similar patients. The posttreatment Hamilton ratings of depression were significantly reduced in the estrogen-treated group, but not in the placebo group. Possible physiological mechanisms are discussed. The risk-benefit ratio for estrogen therapy of depression in these patients was judged to be favorable. However, periodic endometrial biopsies are required to monitor the endometrial response of women receiving high doses of estrogens.