Purpose <p>This study aimed to evaluate whether the 9th edition of the tumor–node–metastasis (TNM) staging system more accurately reflects prognosis than the 8th edition in patients with non-metastatic lung cancer staged clinically using ¹⁸F-fluorodeoxyglucose positron emission tomography/computed tomography (¹⁸F-FDG PET/CT).</p> Material-methods <p>In this retrospective, two-center study, 1,590 patients diagnosed with lung cancer were initially evaluated. After excluding those with distant metastases at diagnosis, 529 patients (mean age: 65.9 ± 8.7 years) with stage I–III non-metastatic disease who underwent ¹⁸F-FDG PET/CT between February 2017 and December 2023 were included. All patients were reclassified according to both the 8th and 9th editions of the TNM system, and changes in staging (upstaging, downstaging, or unchanged) were analyzed in relation to overall survival (OS).</p> Results <p>The 5-year OS rates were 75.3% for cN0, 55.9% for cN1, 50.0% for cN2a, and 33.7% for cN2b. Although cN2a patients had higher OS compared to cN2b, the difference was not statistically significant (<i>p</i> = 0.20). Importantly, patients with T1N2a tumors who were down-staged from stage IIIA to IIB showed significantly lower OS compared to those consistently classified as stage IIB (<i>p</i> = 0.029), suggesting prognostic mismatch. In contrast, survival trends for patients down-staged from IIB to IIA (T1N1) and up-staged from IIIA to IIIB (T2N2b) were consistent with the revised TNM 9 classification, though without statistical significance (<i>p</i> = 0.19 and <i>p</i> = 0.085, respectively).</p> Conclusion <p>While several staging changes in the 9th TNM edition align with prognostic expectations, the downstaging of the T1N2a subgroup appears to be inconsistent with actual survival outcomes. These findings highlight the need for further large-scale, prospective, multicenter studies to validate the clinical reliability and implications of the revised staging system.</p>