Background: The analysis of immune features of tumor microenvironment is leading to the development of new immunotherapies and the identification of new biomarkers. In this study, we have investigated the prognostic value of immune markers, especially those related to immune-regulation in resectable NSCLC. Methods: RNA was isolated from fresh-frozen lung specimens (tumor and normal lung) (n = 178). RT-PCR was performed to analyze the expression of 20 genes: CCL2, CCL22, CD1C, CD127, CD209, CD25, CD4, CD8, CLEC4, CTLA4, FOXP3, IDO1, IL10, IL23A, LGALS1, LGALS2, NRP1, PD1, PDL1 and TGFB1, by the use of hydrolysis probes. Relative gene expression was assessed by Pfaffl formula and normalized by the use of ACTB, GUSB and CDKN1B as endogenous genes. Statistical analyses were considered significant at p < 0.05. Results: Unsupervised hierarchical analysis according to gene expression levels classified patients in two clusters. Cluster I (n = 70) was composed by samples showing lower expression levels of the analyzed genes, whilst cluster Cluster II (n = 84) grouped samples that, in general have higher gene expression levels. Survival analysis showed that patients in Cluster II had better OS (NR vs 46.6 months, p = 0.040) and longer PFS (81.22 vs 26.28 months, p = 0.027) than patients in Cluster I. In order to obtain a more clinically valid biomarker, an expression score based on the regression coefficients from a multivariate model was built: (IL23A x 0.016) + (IL10 x 0.077) + (CCL2 x 0.154) + (PD1 x 0.132) + (CTLA4 x 0.173). Patients with high expression score have longer OS (NR vs 42.9 months, p = 0.007) and longer PFS (82.6 vs 23.4 months, p = 0.011). Multivariate analysis, including all significant variables found in the study, indicated that the expression score, along with KRAS status, was an independent biomarker of prognosis for OS [HR: 0.368; 95%CI, 0.190-0.715; p = 0.003] in our cohort. Conclusions: We identified a unique immune-reaction related profile associated with survival which leaded to the creation of an expression score composed by five genes. Furthermore, this expression score was found to be an independent prognostic biomarker in resectable NSCLC. Supported by grants PI1202838 and RD12/0036/0025 from ISCIII.