The estrogen receptor protein content of recurrent breast cancer correlates well with the clinical response of hormonal manipulation. This predictive value of the ER of the primary tumor obtained at the time of mastectomy has not yet been proven. If this predictive capability should hold for primary tumors and eventual endocrine treatment, the ramifications are obvious: endocrine therapy could be offered to patients on a rational basis and adjuvant therapy could be considered on a plausible biochemical basis. This report details our observations as to the ER content of the primary tumor and the eventual result of endocrine therapy. Thirty-seven patients whose tumor ER was determined from the primary tumor eventually underwent some form of endocrine therapy for recurrent disease. Fifteen of the primary tumors had significant ER content and 22 possessed insignificant amounts. Only one of the 22 patients with insignificant ER content was benefited by endocrine treatment. Those patients whose tumors contained low amounts of ER experienced recurrence of their disease more rapidly than did those with higher ER content. There was no correlation of age, cell type of tumor or metastatic site with the ER content of the tumor. There is a positive correlation between response to chemotherapy and ER content of tumor. Measurement of the estrogen receptor protein content of the primary breast tumor is a reliable method for choosing patients for eventual endocrine therapy. Those patients whose tumors contain insignificant estrophilin are not candidates for such attempts at palliation.