We have examined two new oestrogen receptor (ER) assays--an enzyme immunoassay (EIA) and an immunocytochemical assay (ICA) in a large series of primary breast tumours to compare their potential as predictors of (1) response to endocrine therapy and (2) survival in patients developing advanced breast cancer. Response to endocrine therapy was categorised at 6 months (UICC criteria). ER-ICA appears the better predictor of response to endocrine therapy than ER-EIA. Combining ICA and EIA results did not improve the prediction of response. With both assays patients with ER positive tumours survived longer from the time of diagnosis of advanced disease than those with ER negative tumours. The predictive power of these assay for progression of disease appears slightly better for the ER-ICA.