Whole-body (18)F-deoxyglucose positron emission tomography (FDG-PET) has the potential to improve the management of non-small-cell lung cancer (NSCLC). We prospectively evaluated the impact of combining FDG-PET with conventional staging methods, including computed tomography (CT), on the staging and management of patients with potentially resectable NSCLC.Ninety-four consecutive patients with newly diagnosed/suspected NSCLC were enrolled. Each patient was first staged by using conventional methods, and then by FDG-PET. FDG-PET results were forwarded in a sealed envelope and divulged at the weekly staff meeting on staging and treatment, only after "Decision 1", based on conventional staging, had been reached by consensus; reevaluation taking FDG-PET into account yielded "Decision 2". The validity of these latter decisions was analyzed retrospectively.Eighty-nine patients were eligible. Relative to standard imaging, FDG-PET led to clinical staging changes in 26 (29.2%) patients. The stage was lowered in eight cases (9%) and raised in 18 cases (20.2%). "Decision 2" differed from "Decision 1" in 19 patients, modifying the surgical procedure in four cases, indicating other investigations to confirm FDG-PET evidence of metastases in 12 cases, or modifying the medical treatment in three cases. These modifications were retrospectively justified in 9/19 cases, and consisted of 2/4 modifications of the surgical procedure (one hilar and one adrenal metastasis not confirmed histologically), 4/12 further investigations (axillary and liver biopsies, mediastinoscopy, occult colon cancer) and three indications for palliative treatment, in patients who all died within 3 months after FDG-PET.Based on FDG-PET, management was modified in 19/89 (21.3%) patients, but these changes were justified in only 9/89 patients (10.1%). FDG-PET can detect asymptomatic local and distant metastases and improves the preoperative assessment of NSCLC, thereby avoiding unnecessary surgery. However, histological verification is required because of the risk of false-positive results.