Context .- Although most primary cancers of the lung carry a heavy mutational load and will potentially present many "nonself" antigens to the immune system, there are a wide range of possible mechanisms for tumors to avoid so-called immune surveillance. One such mechanism is the adoption of immune checkpoints to inhibit the host immune response. Immune checkpoint inhibitors show great promise in the treatment of advanced non-small cell lung cancer. Objective . -To discuss the possibility of biomarker selection of patients for these therapies. This is becoming a much debated issue, and the immunohistochemical detection of Programmed Death Ligand 1 (PD-L1), the ligand for the inhibitory Programmed Death receptor 1 (PD-1) checkpoint, is one possible biomarker. Data so far available show some conflicting results, but PD-L1 immunohistochemistry looks likely to be introduced into clinical use for selecting patients for treatment with anti-PD-1 or anti-PD-L1 therapies. Given that there are 4 such drugs rapidly approaching regulatory approval, each with its own independent PD-L1 immunohistochemistry biomarker test, both oncologists and pathologists face some significant challenges. Data Sources .- Peer-reviewed literature and meeting proceedings, especially during the last 12 months, were used. Conclusions . -The biology of PD-1/PD-L1 is complex, the clinical data for these drugs show considerable variation, the selection performance of the PD-L1 biomarker test is not perfect, and the existence of 4 drug/test combinations adds significantly to the problems faced. This article addresses some of the background to this therapeutic problem and discusses some of the issues ahead.