Busulfan, at a dose of 16 mg/kg, is widely used in combination with cyclophosphamide as a conditioning regimen for patients undergoing bone marrow transplantation. Wide interindividual variation in busulfan kinetics and rapid clearance of the drug have been reported, especially in children. Some of the factors contributing to interpatient variability have been identified. They include circadian rhythms, age, disease, drug interaction, changes in hepatic function, and busulfan bioavailability. In this study, we demonstrate that hepatic glutathione S-transferase (GST) activity correlates negatively with busulfan maximum and minimum concentrations (Pearson's correlation r = -0.74 and -0.77, respectively) and positively with busulfan clearance (Pearson's correlation r = 0.728) in children with thalassemia major in the age range of 2 to 15 years. We also found that plasma alpha GST levels were 5 to 10 times higher in patients with thalassemia than in normal controls and age-matched leukemic patients, either reflecting extensive liver damage, elevated expression of the enzyme, or both in thalassemic patients. Plasma alpha GST concentrations showed a similar correlation with busulfan kinetic parameters to that observed for hepatic GST. The status of hepatic GST activity accounts, at least in part, for the observed interindividual variation in busulfan kinetics, while the observed association with plasma alpha GST is difficult to explain at present.