Individualized selection of the most effective adjuvant (or neoadjuvant) chemotherapy for breast cancer based on the molecular characteristics of the tumor could improve the risk:benefit ratio of current therapies. It could also streamline the development of new regimens for those who are unlikely to benefit from existing drugs. It is expected that combinations of markers will be more informative to predict response than any single gene and may yield regimen-specific predictors. Novel molecular analytical tools, particularly transcriptional profiling, provide a method to test this hypothesis. Several small exploratory studies have shown encouraging results. This article reviews recent progress in this field including experience from the breast cancer pharmacogenomic marker discovery program at the Nellie B. Connally Breast Center of the University of Texas M. D. Anderson Cancer Center. This manuscript is based on a presentation that was given during the Presidential Symposium of the annual meeting of the Japanese Breast Cancer Society in 2004.