An immunocytochemical method for oestrogen receptor (ER) analysis of breast carcinomas is described and compared with the dextran-coated charcoal (DCC) assay and a cytochemical method. The method described is superior to the purely cytochemical, especially in terms of sensitivity, demonstration of nuclear receptors and correlation with the DCC assay. The DCC assay is reliable except in the case of poorly cellular tumours: in these the immunocytochemical method is far superior, being independent of tumour cellularity. With this new method, lobular carcinomas are shown to be ER-rich compared with their ductal counterparts, a conclusion currently disputed, largely because methods employed mostly fail to take account of tumour cellularity. A combination of the DCC assay and the immunocytochemical method here described gives the maximum information about the ER-status of a particular carcinoma.