Background: immunotherapy with immune checkpoint inhibitors (ICIs) for solid tumors had significantly improved overall survival (OS). Positive response to PD-1/PD-L1 blockades was observed in the treatment of solidtumors. Breast cancer (BC) patients are no exception. However, the efficacy of immunocheckpoint therapy in BC patients remains poor. A particularly important factor is the lack of studies on the expression patterns of immune checkpoints in BC patients. Method: in this study, the microarray dataset GSE10810 was downloaded from the Gene Expression Database Synthesis (GEO) to analyze the differential expression of BC genes and perform GO and KEGG analysis. Then, we explored theprognostic value of ICs for BC patients by analyzing RNA-seq and mutation data from 657 BC patients from theCancer Genome Atlas (TCGA) database. Next, we analyzed the differences in the expression of relevant immune checkpoints between 1085 BC patients and 291 healthy controls (HC) in the Genotype-Tissue Expression (GTEx) database. Finally, we analyzed the expression levels of IFNγ-R, PD-L1, STAT1, IDO and other genes in MCF-10A, MF-10AT, MCF-7, MDA-MB-231 cell lines.Results: It was found that increased expression of PD-1, PD-L1, STAT1, CTLA-4 was associated with poor OS in BC patients. In addition, co-expression of PD-L1 with PD-1, STAT1or CTLA-4 and co-expression of PD-1 with CTLA-4was related to poor OS. We analyzed associations between the proportionate expression of PD-L1 and PD-1, PD-L1 and STAT1, PD-1 and CTLA-4, PD-1 and LAG3, PD-L1 and CTLA-4 in BC patients, there was significance in correlation in both of the BC patients.The expression of STAT1 in BC patients was compared with that of HC, and it was found that STAT1 was highly expressed in BC patients. Conclusions: our results suggest that transcriptome-based co-expression of STAT1 and PD-L1 is a predictor for poor OS in BC patients, which might provide novel insight into designing combinational targeted therapy for BC.