Controversy exists regarding the relationship of the degree of c-erbB-2 amplification to other prognostic factors in breast cancer. To determine the degree of amplification of c-erbB-2 exactly, a sensitive and quantitative method is required. We have developed a competitive PCR method to quantitatively determine the amplification of the c-erbB-2 oncogene. Using this method, we evaluated DNA from 27 breast cancer tissue specimens and DNA from peripheral blood leukocytes from a normal individual. Regarding the relationship between the degree of c-erbB-2 amplification and clinicopathological factors, we found a greater degree of amplification of the c-erbB-2 oncogene in estrogen receptor- negative or progesterone receptor-negative specimens than in positive ones and in lymph node metastasis-positive specimens than in negative specimens, in stages II, III, and IV of disease compared with stage I disease, and in samples with positive lymphatic vessel invasion than with no lymphatic vessel invasion. Generally, these factors were seen in the group of patients who had a bad prognosis. By univariate analysis and multivariate analysis, reverse correlation was observed between amplification of c-erbB-2 and overall survival. Regarding disease-free survival, these relationships were observed only with univariate analysis in our group of patients.