Triple negative breast cancer (TNBC) is the most aggressive and malignant subtype in breast cancer. Immunotherapy is a currently promising and effective treatment for TNBC, while not all patients are responsive. Therefore, it is necessary to explore novel biomarkers to screen sensitive populations for immunotherapy. All mRNA expression profiles of TNBC from The Cancer Genome Atlas (TCGA) database were clustered into two subgroups by analyzing tumor immune microenvironment (TIME) with single sample gene set enrichment analysis (ssGSEA). A risk score model was constructed based on differently expressed genes (DEGs) identified from two subgroups using Cox and Least Absolute Shrinkage and Selector Operation (LASSO) regression model. And it was validated by Kaplan-Meier analysis and Receiver Operating Characteristic (ROC) analysis in Gene Expression Omnibus (GEO) and the Molecular Taxonomy of Breast Cancer International Consortium (METABRIC) databases. Multiplex immunofluorescence (mIF) and Immunohistochemical (IHC) staining were performed on clinical TNBC tissue samples. The relationship between risk score and immune checkpoint blockades (ICB) related signatures was further investigated, as well as the biological processes were performed by gene set enrichment analysis (GSEA). We obtained three DEGs positively related to prognosis and infiltrating immune cells in TNBC. Our risk score model could be an independent prognostic factor and the low risk group exhibited a prolonged overall survival (OS). Patients in low risk group were more likely to present a higher immune infiltration and stronger response to immunotherapy. GSEA revealed the model was associated with immune-related pathways. We constructed and validated a novel model based on three prognostic genes related to TIME in TNBC. The model contributed a robust signature that could predict the prognosis in TNBC, especially for the efficacy of immunotherapy.