Background: PD-L1 is the only biomarker that has been accepted to identify patients who could potentially benefit from immune checkpoint inhibitors (ICIs) administration, but also PD-L1 negative tumors respond to immunotherapy. However, other biomarkers have been proposed but they are too expensive and actually, there is not a consensus for detection and analytical methods. Thus, there is an urgent need for inexpensive and reproducible tests to better optimize the clinical use of ICIs. Recent findings suggest that host parameters may help to select patients to treat with ICIs: performance status (PS) and some laboratory parameters, such as lactate dehydrogenase (LDH) and blood cells count. Inflammation and inflammatory response have been considered surrogate biomarkers of host immunity. An elevated neutrophil-to-lymphocyte (NLR), myeloid-to-lymphocyte (MLR) and platelet-to-lymphocyte (PLR) ratio have been associated with systemic inflammation and some evidences suggest that these biomarkers could play a role as prognostic factors. Methods: We performed a retrospective analysis of 250 patients treated with ICIs at our institution. The associations between basal NLR, MLR and PLR, clinicopathological features and their impact on progression free survival (PFS), overall survival (OS) and toxicities were assessed. Results: From May 2015 to January 2022, 250 patients were screened, of whom 66 were excluded from the analysis. A total of 184 patients (64.7% men), with a median age of 68 years (37-90) a PS 0 (52.2%), 1 (39.7%) and 2 (8.1%), were enrolled. The primary tumor origin was lung (93 adenocarcinoma, 19 squamous, 14 small cell lung cancer), urothelial (33), melanoma (4) and renal (21). Nivolumab was used in 103 patients, Pembrolizumab in 67 and atezolizumab in 18. At 24 months median follow up there were 131 progression diseases and 122 deaths. Overall, the median value of NLR, PLR and MLR was 3.15 (1-41), 168 (4-797) and 3.5 (0-245), respectively. Lower values of NLR and PLR were associated with a better median PFS (HR for NLR 1.7, 95% CI 1.2-2.4, p = 0.002; HR for PLR 2.2, 95% CI, 1.6 - 3.2, p < 0.0001) and OS (HR for NLR 1.5, 95%CI 1.1-2.2, p = 0.017; HR for PLR 1.9, 95%CI 1.3-2.7 p < 0.0001). MLR was not statistically significant. According to primary tumor origin, renal cancer reported a value significantly under the median of NLR and PLR (p = 0.015 and p = 0.021 respectively). Toxicity and clinico-pathological factors showed a significant association between female sex and both severe fatigue (p = 0.002) and pain (p = 0.018). Pain was also related to younger age (p = 0.009) and to a PS ≥2 (p = 0.003). Conclusions: Our study shows that NLR and PLR represent inexpensive, reproducible and reliable markers that may help identifying patients who can have a greater benefit from immunotherapy.