BACKGROUND: Current nodal classification inadequately predicts outcomes for advanced non-small cell lung cancer (NSCLC) patients. We investigated whether metastatic lymph node characteristics could improve prognostic accuracy. METHODS: We retrospectively analyzed 339 patients with advanced NSCLC who received immunotherapy as first-line treatment. Lymph node imaging was performed using computed tomography (CT), and the X-tile software was employed to determine optimal cutoff values for lymph node size and number. Prognostic factors were assessed using Kaplan-Meier survival curves and multivariate Cox regression analysis. The predictive accuracy of various N-staging was evaluated through time-dependent receiver operating characteristic (ROC) curves. RESULTS: The optimal cutoff values for lymph node size and number were 1.60 cm and 3, respectively. Kaplan-Meier analysis indicated that size, number, and fusion of metastatic lymph nodes were associated with worse overall survival (OS) in advanced NSCLC patients {hazard ratio (HR) [95% confidence interval (CI)]: 2.179 (1.432-3.316), 1.859 (1.226-2.821), and 3.635 (1.796-7.358)}. Multivariate Cox regression analysis identified lymph node size [HR (95% CI): 6.21 (1.19-32.25)] and fusion [HR (95% CI): 3.20 (1.32-7.75)] as independent prognostic factors for OS. Incorporating lymph node size into the conventional N-staging system improved prognostic accuracy, with a 3-year area under the curve (AUC) of 0.651 (95% CI: 0.535-0.767). CONCLUSIONS: Lymph node size serves as a valuable indicator of tumor invasion and can enhance the existing N-staging system for more accurate prognosis prediction for more accurate prognosis prediction in advanced NSCLC.