Although chemotherapy is widely used to treat human cancers, most chemotherapeutic agents only benefit a small fraction of patients because of the heterogeneity of cancers. Therefore, identifying of the sensitivity of cancers toward various chemotherapies would be important for choosing of chemotherapeutic regime. In this study, a 23-gene chemoresistance signature was developed from chemoresistant breast cancers. Functions of the genes in the signature were related with transcription and translation. The signature was indicative of chemoresistance and associated with poor prognosis in multiple chemotherapeutic agents and cancer types. Furthermore, by applying computational approaches, we identified several compounds that might specifically affect the chemoresistant signature. Decitabine (DAC) was the compound most likely to target the signature. In vitro and clinical analysis confirmed effect of DAC toward both breast cancer cell lines and ovarian cancers respectively. In conclusion, our study identified a chemoresistant signature that is both predictive and prognostic, and the signature-related chemoresistance could be suppressed by DAC treatment.