Abstract Background: We previously published a 69-gene DDR microarray signature that predicted for pathologic complete response (pCR) to anthracycline chemotherapy in women with basal-like breast cancer in two clinical studies1. Here, we sought to validate this predictive signature in independent data sets of breast cancers, selecting for basal-like and luminal B subtypes. We also customized a low density qRT-PCR array for use on formalin-fixed paraffin embedded tissue that will allow for prospective validation and clinical use. Methods: Six independent validation sets were selected with known response to DNA damaging agents, containing either anthracyclines or cisplatin2-7. One independent data set with known response to the microtubulin agent docetaxel (non-DNA damaging) was used. 8 Using PAM50 genes, the tumors were first classified according to the molecular subtype. The predictive value of the 69-gene DDR assay was then correlated with known pathologic response, and the receiver operating curve determined. We then performed qRT-PCR of these 69 genes with 5 controls, and correlated the expression of these genes with microarray RNA values in an additional 50 basal tumors. The data for individual genes was normalized, combined, and then hierarchically clustered. Results: In the 170 basal tumors treated with DNA damaging agents, a high score predicted response in basal-like tumors. A, p< 0.05. Conversely, for a non-DNA-damaging therapy, a high score significantly predicted for lack of response. B. In the 116 luminal B cancers, the DDR-assay also predicted for response to DNA damaging agent containing regimens with AUCs ranging from 0.67 to 0.80. C. Microarray of the 69 DDR gene expression values was compared with Ct values by q-RTPCR low density array for 50 paired samples. High concordance between these two platforms (48/50 samples) was observed. Conclusion: We have identified a robust predictive DDR assay that predicts for response to the DNA damaging agents, doxorubicin, epirubicin, and cisplatin in multiple studies, but not for docetaxel therapy. Future studies are underway to test for prediction of benefit of PARP1 inhibitor in basal-like and luminal B breast cancer.