Immune checkpoint inhibitors utilize the immune system to kill cancer cells and are now widely applied across numerous malignancies. Pembrolizumab has two breast-specific indications in triple-negative disease. Currently, programmed death ligand-1 (PD-L1) expression on tumor and surrounding immune cells is the only validated predictive biomarker for immune checkpoint inhibitors (ICIs) in breast cancer; however, it can be imprecise. Additional biomarkers are needed to identify the patient population who will derive the most benefit from these therapies. The tumor immune microenvironment contains many biomarker candidates. In tumor cells, tumor mutational burden has emerged as a robust biomarker across malignancies in general, with higher burden cancers demonstrating improved response, but will need further refinement for less mutated cancers. Preliminary studies suggest that mutations in breast cancer gene 2 (BRCA-2) are associated with increased immune infiltration and response to ICI therapy. Other genomic alterations are also being investigated as potential predictive biomarkers. In immune cells, increased quantity of tumor-infiltrating lymphocytes and CD8+ cytotoxic T cells have correlated with response to immunotherapy treatment. The role of other immune cell phenotypes is being investigated. Peripherally, many liquid-based biomarker strategies such as PD-L1 expression on circulating tumor cells and peripheral immune cell quantification are being studied; however, these strategies require further standardization and refinement prior to large-scale testing. Ultimately, multiple biomarkers utilized together may be needed to best identify the appropriate patients for these treatments.