PurposeRisk stratification of patients with non-small cell lung cancer (NSCLC) is crucial to select the appropriate treatments, but available models for patients with complete resection are unsatisfactory. The purpose of this study was to determine a prediction model based on clinical information, routine physical and blood tests, and molecular markers.Patients and MethodsThis was a retrospective cohort study of patients who underwent surgical resection for lung cancer between 2009 to 2013. Potential prognostic factors were used to build a full prediction model based on a multivariable Cox regression analysis. A nomogram was constructed. The risk stratification cutoffs for clinical use were determined based on the model.ResultsA total of 368 NSCLC patients with R0 resection were included. The final multivariable model indicated that low diffusing capacity of the lung for carbon monoxide (HR=1.66, 95% CI: 1.18–2.34), high platelet-to-lymphocyte ratio (HR=1.42, 95% CI: 1.04–1.95), histology type of squamous cell carcinoma and others (squamous cell carcinoma vs adenocarcinoma, HR=1.40, 95% CI: 1.01–1.96; others vs adenocarcinoma, HR=2.36, 95% CI: 1.15–4.84; P trend=0.001), N>0 status (HR=1.96, 95% CI: 1.42–2.70), high serum carcinoembryonic antigen levels (HR=1.61, 95% CI: 1.13–2.27), and postoperative chemotherapy (HR=0.53, 95% CI: 0.33–0.87) were independently associated with poor OS. The patients were classified into four risk groups according to the nomogram, and the OS was different among the four groups (P<0.05).ConclusionA nomogram was successfully constructed based on a multivariable analysis, and the nomogram can discriminate the OS of patients with NSCLC based on risk categories, but external validation is still necessary.