Background: The purpose of this study was to assess the efficacy of the different techniques of lymph node biopsies in patients with suspected metastatic non-small cell lung cancer (NSCLC) in the subaortic (station #5) and paraaortic (station #6) lymph nodes. Methods: This was a retrospective cohort study conducted of a prospective database of patients between January 2003 and June 2006 with suspected N2 disease only in the #5 or #6 lymph nodes, or both. All patients had integrated 2-deoxy-2-fluoro-d-glucose positron emission tomography/computed tomography, and nodal biopsy or thoracotomy, or both, with complete thoracic lymphadenectomy. Results: There were 112 patients with clinically suspected N2 disease in lymph node stations #5 or #6, or both. The primary tumor was in the left upper lobe in 98 (88%) and in the left lower lobe in 14 (13%), and 58 had pathologic N2 disease in #5 or #6 lymph node stations only. Mediastinoscopy, used in all patients found, unsuspected N3 disease in 4 patients (3.6%) and N2 (#4L) disease in 12 (11%). Endoscopic ultrasound with fine needle aspiration (EUS-FNA), implemented in 62 patients (56%), correctly identified 41 patients (66%). Left single-incision video-assisted thoracic surgery (VATS) was used in 39 patients and was correct in 100%. Of the 58 patients, 53 (91%) completed neoadjuvant chemoradiotherapy, followed by resection, and their 5-year survival was 64%. Conclusions: EUS-FNA is less accurate for the #5 and #6 lymph node stations than left VATS. We prefer left VATS over the Chamberlain procedure for patients with suspected nodal metastases isolated only to #5 or #6 stations, and if positive, we prefer neoadjuvant therapy. The advantage of neoadjuvant therapy followed by resection compared with resection followed by adjuvant therapy remains controversial; and hence, the role for biopsy of these nodes is also controversial. © 2007 The Society of Thoracic Surgeons.