This study reports on our experience with 18F-fluorodeoxyglucose (FDG) positron emission tomography (PET) in pediatric patients affected by Hodgkin's disease (HD) and non-Hodgkin's lymphoma (NHL). We studied 20 pediatric subjects (12 males, 8 females; mean age, 10 years; range, 6 months to 14 years) with malignant lymphoma (9 HD, 11 NHL) for a 4-year period of time. Overall, 45 PET scans were performed: 7 at disease presentation and 38 for evaluation of response to therapy or follow-up study. All PET results were compared with conventional imaging (CI), mainly computed tomography (CT) and/or magnetic resonance imaging (MRI), and supported by clinical follow-up and/or histologic data. In 18 of 20 patients, PET findings correctly identified the status of disease. Two (2) subjects (respectively, 1 HD and 1 NHL, both at follow-up) resulted falsely positive: 1 due to prominent thymic uptake, and the other due to nonspecific inflammation. Of 45 scans, PET findings were consistent with clinical follow-up and other CI data in 43 cases (16 true-positive and 27 true-negative results) and resulted falsely positive in the remaining 2 scans. On a lesion-by-lesion basis (overall, 153 lesions: 84 nodal and 69 extranodal), we found a concordance between CI and PET findings in 25 nodal (29.8%) and in 22 extranodal sites (32%). PET was more accurate than CI, as it identified active disease in 1 patient negative at CI and excluded relapse in 6 patients with inconclusive CI and in 2 patients with a falsely positive CI. Overall, PET sensitivity and specificity was 100% and 93% versus 94% sensitivity and 72.4% specificity for CI. This comparative study shows FDG PET to be more accurate than CI in evaluating children with lymphoma. Our data also confirms that (18)F-FDG PET may show false-positive findings.