An estrogen receptor-immunocytochemical assay (ER-ICA) was performed on frozen sections of 130 samples of human breast carcinoma. A standard dextran-coated charcoal assay (DCCA) was performed on the same samples. Concordance of results between the tests was 91%. The sensitivity and specificity of the ER-ICA, compared with the DCCA, were 92% and 89%, respectively. We describe the ER-ICA technique and review the literature regarding the use of the ER-ICA in evaluating breast cancer with respect to the agreement of results with the DCCA, the nature of discordant results, the ability to predict the clinical response to hormone therapy, and the ability to predict disease-free survival. The combined experience of many studies has shown that the ER-ICA is a highly specific and sensitive method for measuring the level of ERs in breast tumors with a high level of agreement with the DCCA. Early experience has suggested that the ER-ICA can predict the response to hormone therapy and disease-free survival, as well as or better than the DCCA. The evaluation of receptor heterogeneity, made possible by the ER-ICA, may enhance our ability to discriminate ER-positive tumors with a relatively high risk of recurrence.