This study was designed to determine the value of 2-[fluorine-18]-fluoro-2-deoxy- d-glucose positron emission tomography (FDG-PET) in the early assessment of therapy response in lymphoma patients. We studied 20 patients with pathologically proven lymphoma, including 17 patients with aggressive non-Hodgkin's lymphoma and three patients with Hodgkin's lymphoma. All patients underwent whole-body FDG-PET imaging at baseline and after 1-2 cycles of chemotherapy. PET images were analysed visually and quantitatively by calculating the standardised uptake value (SUV). In each patient, we measured the SUV of the tumour demonstrating the highest FDG uptake at baseline study and the SUV of the same tumour after 1-2 cycles of therapy. The achievement of complete response was assessed on the basis of a combination of clinical findings and the results of conventional imaging modalities. Follow-up of progression-free survival (PFS) was obtained for the validation of PET data. Of the 20 patients, ten achieved complete remission at the completion of chemotherapy and the other ten did not respond to chemotherapy. Of the ten responders, four are still in remission (PFS 24-34 months) while the other six have relapsed (PFS 8-16 months). For the prediction of 24-month clinical outcome, visual analysis of PET after 1-2 cycles showed high sensitivity (87.5%) and accuracy (80%) but low specificity (50%). Comparison with the baseline SUVs revealed that the responders showed a significantly greater percent reduction in SUV after 1-2 cycles of therapy as compared with the non-responders (81.2%+/-9.5% vs 35.0%+/-20.2%, P<0.001). In addition, using 60% reduction as a cut-off value, the responders were clearly separated from the non-responders, with the exception of one non-responder. In conclusion, when performed early during chemotherapy, FDG-PET may be predictive of clinical outcome and allows differentiation of responders from non-responders in cases of aggressive lymphoma.