We have determined the predictive value of [18F]2-fluoro-2-deoxy-glucose (FDG-PET) in patients with Hodgkin's disease (HD) and aggressive non-Hodgkin's lymphoma (NHL) scheduled for high-dose therapy with stem cell transplantation (HDT/SCT). Inclusion criteria were the presence of an FDG-PET scan after chemotherapy (ChT) within 8 weeks prior to HDT/SCT and available follow-up data. Sixteen patients (10 NHL and six HD) were observed during a follow-up period of 4 to 28 months (median 13 months). Before SCT, five patients had a negative PET, three were weakly positive, two moderately positive, and six strongly positive. None of the five patients with a negative PET before HDT/SCT relapsed and two of three patients with a weakly positive scan are still in remission after HDT/SCT. Of eight patients with a moderate or high positive PET before HDT/SCT, seven relapsed and one died of early HDT/SCT related complications (P< 0.01). Three of eight relapsing patients died of lymphoma 5 to 10 months after SCT and in one additional patient not responding to HDT/SCT, the main cause of death was chronic toxicity 4 months after transplantation. After 12 months, in PET-negative patients the overall and relapse-free survival was 100%, in PET-positive patients 55% and 18%, respectively. In NHL, two patients with negative PET, but with an age-adjusted international prognostic index (AaIPI) of 2 and one with AaIPI = 1 are still in remission. In the seven PET-positive subjects, one patient with AaIPI = 0, three with AaIPI = 1, and two with AaIPI = 2 relapsed. We conclude that FDG-PET is accurate in the prediction of relapse prior to HDT/SCT in patients with lymphoma. It provides additional information when compared with the AaIPI.