Most of the data regarding the signijcance of c-erbB-2 oncogene expression as a prognostic marker in breast cancer have been generated in many large retrospective studies by retrieving the corresponding oncoprotein in archival parafin em bedded sections. Recent@. employing fresh breast cancer cells obtained by means offine-needle aspiration biopsy, we found a rate of c-erbB-2 positive breast tumors (58%) higher than that reported in pamfinem bedded tissue sections by others studies. The present analysis was undertaken to investigate the impact of routine tissue processing on the preservation of the c-erbB-2 immunoreactivity. This issue was addressed by assessing the relative rate of c-erbB-2 oncoprotein immunodetection on FNAB smears and matched surgical specimens of breast cancer. The expression of c-erbB-2 oncoprotein was evaluated using the alkaline phosphate-anti-alkaline phosphatuse (APAAP) technique in 54 breast aspirates and corresponding surgical specimens ofprimary breast cancer. Twenty-six (48%) smears and 23 (43%) matched parafin sections gave specifc signal for c-erbd-2 oncoprotein. The slightly higher incidence of c-erbB-2 expression found on smears seems to be mainly due to the better antigen preservation in the fresh cytologicalpreparations. We conclude that routine histological processing may afect c-erbB-2 immunoreactivity; therefore, in mounting prospective studies, it is advisable to assess c-erbB-2 status in fresh tissue. Moreover, the assessment of c-erbB-2 expression on aspimte samples may yield additional information to the pre-surgicalprognostic evaluation of breast cancer diagnosed by FNAB. Diagn