[en] Purpose: The aim of this study was to retrospectively evaluate the performance of positron emission tomography (PET) with F-18-fluorodeoxyglucose (F-18-FDG) in children with lymphomas, at various stages of their disease. Methods: Twenty-eight children (mean age 12.5 years, 14 girls, 14 boys) with Hodgkin's disease (HD, n=17) or non-Hodgkin's lymphoma (NHL, n= 11) were evaluated. Patients were investigated at initial staging (n=19), early in the course of treatment (n=19), at the end of treatment (n=16) and during long-term follow-up (n=19). A total of 113 whole-body PET studies were performed on dedicated scanners. PET results were compared with the results of conventional methods (CMs) such as physical examination, laboratory studies, chest X-rays, computed tomography, magnetic resonance imaging, ultrasonography and bone scan when available. Results: At initial evaluation (group 1), PET changed the disease stage and treatment in 10.5% of the cases. In early evaluation of the response to treatment (group 2), PET failed to predict two relapses and one incomplete response to treatment. In this group, however, PET did not show any false positive results. There were only 4/75 false positive results for PET among patients studied at the end of treatment (group 3, specificity 94%) or during the systematic follow-up (group 4, specificity 95%), as compared with 27/75 for CMs (specificity 54% and 66%, respectively). Conclusion: F-18-FDG-PET is a useful tool for evaluating children with lymphomas. Large prospective studies are needed to appreciate its real impact on patient management.