Estrogen receptor assays were performed in 13 cases of carcinoma of the breast on two samples of tumor tissue. One was the usual sample obtained from the specimen submitted for frozen section. The other sample was taken from tumor tissue remaining in the mastectomy specimen. Eight of the 13 cases were positive for estrogen receptors determined by the dextran-coated charcoal method. The initial frozen section specimens showed estrogen receptor levels 1.6 to 6.5 times greater than the mastectomy specimens, which were lower in seven of eight cases. One case showed no change in estrogen receptor levels. The devitalization of the breast during mastectomy, mechanical and possible chemical injury to the tumor, and delays in obtaining the mastectomy specimen are factors that cause a decreaSE in receptor levels. A tumor with a low level of estrogen receptor might possibly yield a false negative result in a mastectomy specimen.