A new immunocytochemical assay (ICA) for progesterone receptor (PgR), employing the rat monoclonal KD-68 antibody and a sensitive peroxidase-anti-peroxidase (PAP) technique as the displaying system, was performed in 129 human breast cancer specimens. PgR-ICA staining was almost all electively located in neoplastic cell nuclei with a substantial heterogeneity in distribution and intensity. To study the basic relationship of the results of the ICA method with the biochemical dextran-coated charcoal (DCC) assay we compared, in all the same specimens, the antibody nuclear staining with the PgR positivity by DCC (cut-off value of 10 fmol/mg of protein). We found an overall agreement of 77% between the two methods and a PgR-ICA sensitivity of 83% and a specificity of 72%, assuming that biochemical PgR is truth. PgR-ICA false-negative results were only nine out of 53 (17%); and false-positive were 21 out of 76 (28%). Using both methods no significant association was observed between PgR positivity with menopausal status, histological type, tumor size and lymph node status. The correlations between PgR expression and cell kinetics were assessed by an immunocytochemical method employing the monoclonal Ki-67 antibody. While a significant negative relationship was found between high Ki-67 score and PgR-ICA positivity (P less than 0.01) no correlation was found with DCC positivity. The present results demonstrate that ICA is a practical, reliable and inexpensive method with a good correlation to the conventional biochemical assay to determine the PgR status. Moreover, ICA recognizes PgR expression at the single cell level, thus providing additional information to the quantitative DCC assay that should improve the prognostic evaluation and the prediction of responsiveness to endocrine therapy in breast cancer.