Background: Inflammatory cells have important effects on tumor development. Systemicinflammation markers can be used as prognostic factors. Numerous studiesshown that high pretreatment neutrophil-to-lymphocyte ratio (NLR) and/or platelet-to-lymphocyte ratio (PLR) levels are potential prognostic predictors for poor progression-free survival (PFS) and overall survival (OS) in NSCLC patients receiving immunotherapy. Methods: We performed a cohort study of patients with metastatic or recurrent NSCLC treated with nivolumab monotherapy in second‐line or further‐line treatment in Clinical hospital centre Zagreb. Pre-treatment NLR and PLR were calculated by division of neutrophils and platelets by lymphocytes measured in peripheral blood. Patients were categorized in two sub-groups according to their NLR and PLR values. In previous meta-analyses it was suggested that significant cut-off value of NLR is NLR < 5 and ≥5 and PLR < 160 and ≥160. We analysed PFS and OS. Results: Overall 105 patients diagnosed with NSCLC were treated with nivolumab. The patients were enrolled from March 2017 until October 2017 and were observed them for disease progression and death until June 1st 2020. Most of the patients were male (71; 67.6%) with median age 60.3 years (36-77). Our patients were selected on the basis of good performance status, so most of them had ECOG PS 0 and 1 (103; 98.1%). Therapy was applied mostly in the second and third line (67; 64%), but even up to seventh line (2; 1.9%). Median duration of therapy was 34.5 weeks (2-149), while median number of doses was 17 (1-69).The median PFS was 7.2 months (95% CI 4.53-9.86). Regardless of previous treatment the mOS was 16.1 months (95% CI 11.26-20.93).We observed median value of NLR 4.08 (IQR 2.44-5.84) and PLR 200 (IQR 127.49-284.72). Patients with low PLR had better overall survival compared to patients with low PLR (mOS 20.5 months vs 11.9 months; 95%CI 14.07-26.92 vs 7.35-16.44; p = 0.039). The same was not as clear in mPFS, tendency of better mPFS was toward low PLR, but it did not reach statistical difference (low PLR mPFS 9.1 months vs high PLR mPFS 6.1 months; p = 0.49).Patients with low NLR had significantly better overall survival compared to patients with high NLR (mOS low NLR 18.2 vs high NLR 10.1 months; 95%CI 13.07-23.32 vs 6.04-14.15; p = 0.014). Again, the statistical significance was not reached for progression-free survival (mPFS low NLR 8.3 months vs high NLR 5.8 months; 95%CI 4.81-11.78 vs 2.91-8.68; p = 0.214). Conclusions: Here, we demonstrated that the presence of indicators of systemic inflammation suchas high NLR and high PLR are associated with poor overall survival, but not withprogression-free survival in pre‐treated NSCLC patients who received nivolumabtreatment. The limitation of our study is the lack of a randomizedcontrol and small sample size. The main strength of our study is that it is real-worldeveryday clinical setting.