Lung cancer is the most common cause of cancer death with unchanged mortality for 50 years. Only localized nonsmall-cell lung cancer (NSCLC) is curable. In these patients it is essential to accurately predict survival to help identify those that will benefit from treatment and those at risk of relapse. Despite needing this clinical information, prospective data are lacking. We therefore prospectively identified prognostic factors in patients with potentially curable lung cancer. Over 2 years, 110 consecutive patients with confirmed localized NSCLC (stages 1-3A) were recruited from a single tertiary center. Prognostic factors investigated included age, gender, body mass index (BMI), performance status, comorbidity, disease stage, quality of life, and respiratory physiology. Patients were followed up for 3-5 years and mortality recorded. The data were analyzed using survival analysis methods. Twenty-eight patients died within 1 year, 15 patients died within 2 years, and 11 patients died within 3 years postsurgery. Kaplan-Meier survival estimates show a survival rate of 51% at 3 years. Factors significantly (p < 0.05) associated with poor overall survival were age at assessment, diabetes, serum albumin, peak VO(2) max, shuttle walk distance, and predicted postoperative transfer factor. In multiple-variable survival models, the strongest predictors of survival overall were diabetes and shuttle walk distance. The results show that potentially curable lung cancer patients should not be discriminated against with respect to weight and smoking history. Careful attention is required when managing patients with diabetes. Respiratory physiologic measurements were of limited value in predicting long-term survival after lung cancer surgery.