In the present study, using series of bioinformatics techniques, including single-sample gene set enrichment analysis (ssGSEA), ESTIMATE, CIBERSORT, univariate and multivariate Cox proportional hazard regression analysis, and machine learning, we integrated the immune-related signatures with metastasis characteristics and established a metastatic and immunogenomic risk score (MIRS) for breast cancer patients based on a panel of 12 genes. Subsequent analyses indicate that MIRS is associated closely with the overall survival outcomes and is of guiding significance in clinical therapy for patients with breast cancer. Moreover, compared to existed prognosis signature, MIRS exhibits a great performance of prognostic prediction in immunotherapy. Particularly, our analysis showed that MIRS is capable of predicting response to immune checkpoint inhibitors, implying MIRS is of certain value in immunotherapy for breast cancer. For more instructions, please refer to guidlines page.