Anthracyclines containing regimen are widely used in advanced breast cancer. The response to first line chemotherapy varies according to many individual factors and the theoretical response to a given protocol cannot predict the response of a patient. A randomized clinical trial (ERASME) was initiated in order to evaluate the more appropriate first line chemotherapy scheme in advanced breast cancer. Prognostic factors were included in a multiple logistic regression to explain the response after the first 3 chemotherapy courses (monthly FEC). Three factors were found to be statistically significant: adjuvant hormonotherapy, loco-regional metastases, adjuvant adriamycin containing regimen (pejorative prognostic factor). By combining these factors, this statistical model enables us to predict a response rate to a first line chemotherapy from 27 to 87%. Such a model can be taken into account in a decision-making procedure of first line chemotherapy in advanced breast cancer.