Immunotherapy in lung cancer treatment is a long history paved with failures and some successes. During the last decade, the discovery of checkpoints inhibitors led to major advances in treating advanced and metastatic non-small cell lung cancer (NSCLC). Impressive data from early phase I-II studies were subsequently confirmed in large prospective randomized trials and meta-analyses (High-level of evidence). Three anti-programmed death-1 (PD1) (pembrolizumab, nivolumab) or antiPD-ligand(L)1 (atezolizumab) antibodies showed clinically significant improved survival compared to second-line docetaxel. Then, first-line pembrolizumab monotherapy demonstrated its superiority over platinum-doublet in high PD-L1 NSCLC. The addition of pembrolizumab or atezolizumab to chemotherapy derived the same results regardless of the PD-L1 status. On the opposite, antiCTLA4 (Cytotoxic T-Lymphocyte Associated 4) results are currently disappointing in unselected patients while recent development suggest that the combination of antiPD1 and antiCTLA4 (nivolumab-ipilimumab) positively impact on overall survival. Some secondary analyses also showed that immunotherapy has a positive impact on quality of life and that the clinical improvement can be done at an acceptable incremental cost per QALY. A lot of questions remain unresolved: which is the best treatment duration and is it the same for all patients, how to choose the patients that will have the highest benefit of immunotherapy, how to identify the patients who will have rapid progression, how to improve the current data (new targets, new combinations)…