BACKGROUND: A study was undertaken to investigate the accuracy of positron emission tomography (PET) with 2-[18F]-fluoro-2-deoxy-D-glucose (FDG) in the thoracic lymph node staging of non-small cell lung cancer (NSCLC). METHODS: Forty six patients with focal pulmonary tumours who underwent preoperative computed tomographic (CT) and FDG-PET scanning were evaluated retrospectively. Thirty two patients had NSCLC and 14 patients had a benign process. The final diagnosis was established by means of histopathological examination at thoracotomy, and the nodal classification in patients with lung cancer was performed by thorough dissection of the mediastinal nodes at surgery. RESULTS: FDG-PET was 80% sensitive, 100% specific, and 87.5% accurate in staging thoracic lymph nodes in patients with NSCLC, whereas CT scanning was 50% sensitive, 75% specific, and 59.4% accurate. The absence of lymph node tumour involvement was identified by FDG-PET in all 12 patients with NO disease compared with nine by CT scanning. Lymph node metastases were correctly detected by FDG-PET in three of five patients with N1 disease compared with two by CT scanning, in nine of 11 with N2 disease compared with six by CT scanning, an in all four with N3 nodes compared with two by CT scanning. CONCLUSIONS: FDG-PET provides a new and effective method for staging thoracic lymph nodes in patients with lung cancer and is superior to CT scanning in the assessment of hilar and mediastinal nodal metastases. With regard to resectability, FDG-PET could differentiate reliably between patients with N1/N2 disease and those with unresectable N3 disease.