Recent studies have reported that the pharmacokinetics of high-dose busulfan in bone marrow transplantation (BMT) are age-dependent: with the usual dosage of 16 mg/kg over 4 days, systemic exposure is two to four times lower in children than in adults. Data suggested that the dose of busulfan should rather be calculated on the basis of the body surface area (BSA). We measured plasma pharmacokinetics of busulfan in 27 children (mean age, 5.4 years) who were administered a new dosage of 600 mg/m2 over 4 days, ie, 17.8 to 29.2 mg/kg (mean, 24.8 mg/kg), using a gas chromatography-mass spectrometry assay. Our results demonstrate that, with this new dosage, systemic exposure is significantly increased in children compared with that achieved with the usual dosage of 16 mg/kg (6,404 +/- 2,378 v 3,918 +/- 1,170 ng.h/mL; P = .003). Moreover, there is no longer a significant difference in systemic exposure between children treated with this new dosage and adults given a dose of 16 mg/kg of busulfan. However, despite the use of a dosage normalized to the BSA, there is still a wide interindividual variation in systemic exposure, ranging from 3,566 to 13,129 ng.h/mL, which may account for the high incidence of venoocclusive disease (VOD) of the liver that we have already reported. The optimal dosage and schedule of busulfan in children requires a more individual approach that could be based on dose adjustment and plasma level monitoring.