Background: Due to the widespread use of fine needle aspirate biopsy the practice of determining estrogen (ER) and progesterone (PR) receptors in breast carcinoma from cytological smears (CS) is becoming very common. The aim of this study was to determine concordance between ER and PR assessed by immunocytochemical assay (ICA) on CS and FS both evaluated by image analysis since we have found no data in literature on this. Methods: 104 breast carcinoma cases were selected. For all cases ER and PR determination was performed on CS, obtained by light scraping of the freshly cut surface of the excised surgical tumors at the time of frozen section diagnosis, and FS using the same monoclonal antibodies. Computer-assisted image analysis was performed in all cases using CAS 200. Results were expressed as percent positive area of neoplastic nuclei compared with total nuclear area of the examined neoplastic cells. Results: Good correlation was demonstrated between percent positive nuclear neoplastic area by ER-ICA on CS and FS (r = 0.759; P < 0.0001). Concordance of results was 90.19% (P < 0.001). Good correlation was also demonstrated between percent positive nuclear neoplastic area by PR-ICA in CS and FS (r = 0.889; P < 0.0001). Concordance of results was 97.02% (P < 0.0001). Conclusions: Our data suggest that ICA on CS with automated image analysis is efficient in evaluating ER and PR content in human breast cancer, especially when CS is the only method pathologists have to evaluate receptor status e.g. in advanced breast cancer cases when neoadjuvant therapy is necessary before surgery or when surgery is impossible. Cytometry (Comm. Clin. Cytometry) 38:301–305, 1999. © 1999 Wiley-Liss, Inc.