ObjectiveTo explore the immune biomarkers for prognosis of breast cancer and to construct a risk assessment model.MethodsThe gene expression of breast cancer samples was retrieved from The Cancer Genome Map (TCGA) database and immune related genes (IRGs) were retrieved from the ImmPort database. Cox proportional hazards regression and least absolute shrinkage and selection operator (LASSO) regression were used for prognostic analysis. Gene set enrichment analysis ( GSEA) was used to explore biological signaling pathways. ESTIMATE and CIBERSORT algorithms were used to explore the relationship between risk score and tumor immune microenvironment.ResultsNine kinds of immune-related differentially expressed genes independently related to prognosis were identified: adrenoceptor beta 1 (ADRB1), interleukin 12B (IL12B), syndecan 1 (SDC1), thymic stromal lymphopoietin (TSLP), fibroblast growth factor 19 (FGF19), fatty acid binding protein 7 (FABP7), interferon epsilon (IFNE), tumor necrosis factor receptor superfamily member 18 (TNFRSF18) and interleukin 27 (IL27). The risk assessment equation constructed by these nine kinds of genes had powerful predictive ability. The “neurotrophin signaling pathway” and “adipocyte factor signaling pathway” were activated in patients of high-risk group, and “leukocyte transendothelial migration” “WNT signaling pathway” “FcεRI signaling pathway” “valine, leucine and isoleucine biosynthesis” and “protein export pathway” were activated in patients of low-risk group. A variety of tumor-killing immune cells were significantly enriched in the tumor-infiltrating immune cells of patients in the low-risk group. The immunosuppressive immune cells were significantly enriched in tumor infiltrating immune cells of patients in high-risk group.ConclusionIRGs prognostic signatures are an effective potential predictive classifier in breast cancer treatment.