The purpose of this retrospective study was to evaluate the accuracy of positron emission tomography (PET) using 18-F-fluorodeoxyglucose (FDG) in predicting lymphomatous involvement in the hilar and mediastinal regions in the staging and follow-up of patients with malignant lymphoma. One hundred forty-seven thoracic PET studies in 89 consecutive lymphoma patients were reviewed. Static FDG-PET imaging was performed following application of 270 MBq FDG (mean). Results of FDG-PET were compared with the findings of computed tomography (CT) in all patients and clinical follow-up examination. Eighty-nine of 147 (60%) PET studies showed no FDG uptake in the hilar or mediastinal regions, while 58 (40%) studies did detect FDG uptake in these regions. In 52 of 58 abnormal studies (90%), lymphomatous involvement of the hilar and/or mediastinal regions seen by CT was present. In the remaining six abnormal PET studies (10%), FDG uptake was considered as false-positive because of missing lesions on corresponding CT scans. In four patients false-positive FDG uptake was observed before treatment, in two patients after completion of therapy. In these two patients FDG uptake after therapy was caused by thymus hyperplasia. The remaining four cases before treatment remained unresolved. Sensitivity of FDG-PET was 96%, specificity 94%, positive predictive value 90%, and negative predictive value 98%, respectively. The present study suggests that FDG-PET has potential value in predicting lymphomatous involvement in the hilar and mediastinal regions. FDG-PET may obviate invasive diagnostic procedures in patients with lymphoma.