Immunohistochemical evaluation is, at present, the most reliable method to evaluate estrogen receptors (ERs) in breast cancer samples. Various quantitative 'scoring' systems have been developed to evaluate both the percentage of ER-positive nuclei and the staining intensity. However, such methods show a high interobserver variability. Densitometric evaluation of ERs provides objective results; however, it is scarcely reproducible because results are expressed in arbitrary units. Reproducibly to quantitate, as fmol/mg protein, the cell content of ERs detected by immunohistochemistry in frozen sections of breast cancer, we set up a densitometric method by using the MCF-7 cell cultures as standard. MCF-7 cell cultures were stimulated with insulin and insulin-like growth factor (IGF-1) to express variable concentrations of ERs. To reveal ERs with the same antibody clone, the immunoenzymatic study was carried out by using the ER-EIA method, whereas immunohistochemical reactions were based on ER-immunocytochemical assay (ICA) method. A standard curve, obtained by plotting the ER fmol/mg protein (by the enzymatic assay in MCF-7 cells) versus the optical-density values (by analyzing the immunohistochemical stain on MCF-7 cytospins), was the reference of our method. This method was validated on four cases of human infiltrating breast carcinoma, and no statistically significant differences were observed between ER concentration measured with the EIA method and that obtained by immunohistochemistry.