The amplification of c-erbB-2 oncogene has been reported to have clinical relevance as a prognostic index in breast cancer. However, controversies still remain about its interpretation, mainly due to the inaccuracy of methods used for this purpose and to the unpredictable variability of the ratio between cancer and normal cells. Accurate quantitative assay, combined with strategies for selection or enrichment of tumor cell populations, could shed a new light on the relationships between molecular alterations and their clinical relevance. In this study, amplification of c-erbB-2 was measured by competitive PCR in 21 aneuploid breast cancers using a multiple DNA competitor both in whole homogenized cancer cells and in aneuploid enriched clones obtained after flow cytometry cell sorting. Most breast cancers (10/12) carrying c-erbB-2 oncogene amplification showed a significant increase in copy number in sorted aneuploid clones, and 2/9 apparently not amplified in basal samples were found to be amplified after being sorted for the aneuploid population. A general concordance between amplification and c-erbB-2 overexpression was found. The mean degree of amplification in sorted aneuploid clones is increased in breast cancers with the highest levels of c-erbB-2 protein overexpression. These data indicate that in breast cancers the amplification of c-erbB-2 oncogene is mainly associated with aneuploid cells.