Objective To investigate the clinical value of prognostic nutritional index PNI for predicting the survival outcome of non-small cell lung cancer NSCLC patients. Method The clinicopathological information and survival data of 150 NSCLC patients who underwent surgical resection in Haian Hospital from January 2017 to January 2019 were retrospectively collected and the PNI value calculated as peripheral blood lymphocyte count× 5+ serum albumin level. Receiver operator characteristic ROC curve was used to determine the optimal cut-off value of PNI. Kaplan-Meier curves were plotted and multivariate Cox regression analysis were conducted to determine the prognostic significance of PNI for NSCLC patients. Result ROC curve showed that the optimal cut-off value of preoperative PNI for predicting overall survival OS of NSCLC patients was 42. 6 and the area under curve AUC value sensitivity and specificity was 0. 766 95%CI = 0. 605-0. 928 73. 3% and 77. 8% respectively. Based on the threshold of preoperative PNI all patients were divided into high PNI ≥42. 6 n = 108 and low PNI group PNI<42. 6 n = 42 . By comparing the general information and clinicopathological features between the two groups we found that preoperative low PNI <42. 6 was significantly correlated with advanced age P= 0. 011 ECOG performance status P= 0. 041 and T stage P<0. 001 . Kaplan-Meier curves indicated that the 3-year overall survival OS of patients with PNI< 42. 6 and PNI≥42. 6 were 65. 5% and 96. 3% respectively χ 2 = 21. 922 P< 0. 001 . The preoperative low PNI was a predictor of poor OS in NSCLC patients. Furthermore the results of the multivariate Cox regression analysis demonstrated that ECOG performance status HR = 4. 192 95% CI = 1. 136- 15. 465 P = 0. 031 T stage HR = 6. 832 95%CI = 2. 014-23. 178 P= 0. 002 lymph node metastasis HR= 2. 836 95%CI = 1. 001-8. 038 P= 0. 048 and preoperative PNI HR= 5. 069 95%CI = 1. 330-19. 317 P = 0. 017 were independent prognostic factors for NSCLC patients. Conclusion Preoperative PNI might be used as a reliable predictor of OS in NSCLC patients with a significant clinical value.