Programmed death 1 (PD-1, CD279) and programmed death ligand 1 (PD-L1, CD274) are involved in generating tumor-associated immunosuppression by suppression of T-cell proliferation and interleukin 2 (IL-2) production and immune checkpoint inhibitors targeting these molecules are showing compelling activity against a variety of human cancers. PD-L1 expression has shown a positive association with response to PD-1 inhibition in noncentral nervous system (CNS) tumors, e.g., melanoma or non-small cell lung cancer, and is discussed as a potential predictive biomarker for patient selection in these tumor types. This review summarizes current knowledge and potential clinical implications of PD-L1 expression in glioblastoma. At present, the following conclusions are drawn: (a) functional data support a role for PD-1/PD-L1 in tumor-associated immunosuppression in glioblastoma; (b) the incidence of PD-L1-expressing glioblastomas seems to be relatively high in comparison to other tumor types, however, the reported rates of glioblastomas with PD-L1 protein expression vary and range from 61 to 88%; (c) there is considerable variability in the methodology of PD-L1 assessment in glioblastoma across studies with heterogeneity in utilized antibodies, tissue sampling strategies, immunohistochemical staining protocols, cut-off definitions, and evaluated staining patterns; (d) there are conflicting data on the prognostic role and so far no data on the predictive role of PD-L1 gene and protein expression in glioblastoma. In summary, the ongoing clinical studies evaluating the activity of PD-1/PD-L1 inhibitors in glioblastoma need to be complemented with well designed and stringently executed studies to understand the influence of PD-1/PD-L1 expression on therapy response or failure and to develop robust means of PD-L1 assessment for meaningful biomarker development.