Immune checkpoint inhibitors (ICIs) have been established as a novel strategy for non-small cell lung cancer (NSCLC) therapy. However, a definitive biomarker that can predict response to ICI therapy remains unestablished. The prognostic nutritional index (PNI) is used to assess immune-nutritional conditions and is a prognostic factor in patients with various malignancies; however, its usefulness as a biomarker of response to ICI therapy and survival outcomes in NSCLC patients is unknown. Thus, we retrospectively analyzed the clinicopathological features of advanced-stage or recurrent NSCLC patients treated with ICI therapy to identify predictors of response to ICI therapy and investigate the effects of pretreatment PNI levels on survival after ICI therapy.We selected 102 consecutive NSCLC patients who were treated with ICI therapy from November 2015 to February 2019. We measured their pretreatment PNI levels and performed univariate and multivariate Cox regression analyses of progression-free survival (PFS) or overall survival (OS) after ICI therapy.Pretreatment PNI levels were significantly associated with response to ICI therapy (objective response rate:P = 0.0131; disease control rate: P = 0.0002), PFS (P = 0.0013), and OS (P = 0.0053). In univariate and multivariate analyses of the associations between PNI, C-reactive protein (CRP) or neutrophil-lymphocyte ratio (NLR) and PFS or OS, NLR and PNI, but not CRP, are independent prognostic factors for PFS (NLR: relative risk [RR]=1.655, 95% confidence interval [CI]: 1.012-2.743, P = 0.0449, PNI: RR=1.704, 95% CI: 1.039-2.828, P = 0.0346). Only PNI showed a trend towards being an independent prognostic factor for OS (RR=1.606, 95% CI: 0.952-2.745, P = 0.0761).The pretreatment PNI has the potential to be a simple and novel predictive biomarker of ICI response in NSCLC patients and might help to identify patients who will obtain a survival benefit from ICI therapy.