Targeted therapies or immune checkpoint inhibitors may be the adequate treatment in some patients with advanced non-small cell lung cancer (NSCLC). So far, certain biomarkers are known which may predict tumor response to these drugs. Of major importance is the detection of activating epidermal growth factor receptor (EGFR) mutations. They occur more frequently in the Asian compared to the Caucasian population and are usually found within exon 18-20 of the EGFR gene. Most of them are either a deletion at exon 19 or the L858R point mutation at exon 21. Activating EGFR mutations are predominantly detected in female patients with adenocarcinoma histology and never (or low) smoking history. The efficacy of first (erlotinib, gefitinib) and second (afatinib) generation EGFR tyrosine kinase inhibitors (TKI) in patients with advanced NSCLC and an activating EGFR mutation was demonstrated in several clinical studies. However, at some time during treatment with a first- or second generation TKI usually resistance to these drugs occurs which is mediated by the EGFR T790M mutation in about 50%. Recent trials demonstrated that third generation EGFR TKIs like osimertinib and olmutinib may be effective in T790M mutation positive patients with an overall objective tumor response rate of about 60%. Another molecular aberration which is of importance for treatment decision in patients with advanced NSCLC is the rearrangement of the anaplastic lymphoma kinase (ALK) gene. It was demonstrated that patients with an ALK rearrangement do benefit from therapy with the ALK inhibitor crizotinib. Second generation ALK inhibitors (e.g. alectinib, brigatinib, ceritinib) can overcome resistance which may be mediated by secondary ALK mutations. Moreover, third generation ALK inhibitors (e.g. lorlatinib) were developed and are currently being tested. Similar to patients with an ALK rearrangement also patients with a ROS1 rearrangement may benefit from treatment with crizotinib. The relevance of other molecular characteristics like KRAS or BRAF mutations, c-met amplification and HER2 abnormalities as potential biomarkers for targeted therapies is currently under investigation. Programmed death 1 (PD-1) immune checkpoint inhibitor antibodies like nivolumab or pembrolizumab are used in the clinical routine, however, only about 20% of patients do benefit from this treatment. To use resources as effective as possible, a biomarker to predict tumor response to these type of drugs would be enormous helpful in order to save costs. So far, the impact of expression of the PD ligand 1 (PD-L1) regarding clinical benefit to PD1 and PD-L1 inhibitor therapy was, besides efficacy of these drugs in comparison to chemotherapy, investigated in several randomized clinical trials. While in the CheckMate 017 study the overall survival of squamous cell carcinoma patients treated with nivolumab was independent from PD-L1 expression on tumor cells, in the CheckMate 057 study there seems to be some association between the PD-L1 expression level and the overall survival of adenocarcinoma patients treated with nivolumab. In the KEYNOTE-010 trial especially patients (both squamous cell and adenocarcinoma histology) with high (≥ 50% score) PD-L1 expression on tumor cells did benefit from treatment with pembrolizumab. PD-L1 expression on tumor cells as well as on tumor infiltrating immune cells were investigated in patients treated with the PD-L1 inhibitor atezolizumab in the POPLAR study. The overall survival benefit from atezolizumab increased with increasing PD-L1 expression on tumor cells, tumor infiltrating immune cells or both. Overall, currently the impact of PD-L1 expression and the use of certain cut-off levels to predict response to PD-1 and PD-L1 inhibitors is still under discussion. Different findings may, at least partly, be explained by the use of variables regarding tissue fixation and storage as well as by different antibodies for detection of PD-L1. Alternative biomarker approaches are currently being investigated. In particular, the mutational load as well as the number of predicted neoantigens may be of importance. An association between a higher nonsynonymous mutation burden in tumors and an improved objective response, durable clinical benefit as well as progression-free survival in patients treated with pembrolizumab was reported and, in addition, the efficacy was associated with the molecular smoking signature, higher neoantigen burden and DNA repair pathway mutations. Overall, additional studies are necessary to definitely define the impact of PD-L1 expression as biomarker for PD-1 and PD-L1 inhibitors as well as to investigate the value of alternative biomarkers. In conclusion, strong biomarkers are able to predict response to certain therapies. Thus, ineffective treatment strategies may be prevented and resources including costs may be saved. So far, a few biomarkers are known which are very well established in the clinical routine and are important for treatment decisions in NSCLC patients. Regarding immunotherapy, it seems that the expression of PD-L1 may has some impact to predict response to PD-1 and PD-L1 inhibitors, however, its role needs to be completely clarified. Several candidate biomarkers exist, however, their impact needs to be further investigated. EGFR, ALK, ros1, PD-L1