Background: Positron emission tomography (PET) scans often help direct biopsies of mediastinal lymph nodes in patients with non-small cell lung cancer (NSCLC), but the maximum standard uptake value (maxSUV) of individual nodes has not been evaluated. Methods: This is a prospective study of consecutive patients with NSCLC, all of whom underwent integrated fluorodeoxyglucose-positron emission-computed tomography (FDG-PET-CT) and had biopsy or resection of their mediastinal lymph nodes. Results: There were 397 patients. One-hundred and forty-three patients had N2 disease and 1,252 N2 nodes were pathologically examined. The median maxSUV of the nodes that had metastatic disease were the following: for the 2R node, 10.4 (range, 0-18.6); for 4R, 8.6 (range, 0-18.3); for 5, 8.9 (range, 0-26.3); for 6, 7.6 (range, 0-19.6); for 7, 7.7 (range, 0-14); for 8 and 9, 5.4 (range, 0-8.9). The median maxSUV for all of the N2 nodes that were benign was 0 (range, 0-18.8) (p < 0.05 for all stations except for nodes 8 and 9). When a maxSUV of 5.3 is used the accuracy of integrated FDG-PET-CT for each N2 nodal station is maximized and is at least 92% for each. Conclusions: The maxSUV of individual mediastinal lymph nodes is a predictor of malignancy. There is overlap between false and true positives. Definitive biopsies are required to prove cancer irrespective of the maxSUV value. However, when a maxSUV of 5.3 is used instead of the traditional value of 2.5, the accuracy for FDG-PET-CT for each N2 nodal station increases to at least 92%. © 2006 The Society of Thoracic Surgeons.