In order to evaluate the reproducibility of the interpretation of DNA histograms obtained by two Static Image Analysis Systems (SIAS), the imprints of 50 cases of breast cancers of patients registered at Istituto Nazionale Tumori of Milan were analysed. The imprints were obtained by a single operator. After fixation, the cytological preparations were stained with the Feulgen reaction. The histograms were independently classified, according to Auer's classification, by four investigators. To evaluate the reproducibility of the classification (within investigator agreement) the histograms were submitted to each investigator at two distinct times and the within investigator agreement was estimated by the weighted kappa statistic (kw). The results of the reproducibility of Auer's classification were not entirely satisfactory (kw about 0.70). This preliminary result confirm the suggestion to adopt alternative methods of classification of the ploidy histograms.