We investigated whether adjusting the oral busulfan (BU) dosage on the basis of early pharmacokinetic data to achieve a targeted drug exposure could reduce transplant-related complications in children with advanced hematologic malignancies. Twenty-five children received a preparative regimen consisting of thiotepa (250 mg/m2 i.v. daily for 3 days), BU (40 mg/m2 per dose p.o. every 6 h for 12 doses), and cyclophosphamide (60 mg/kg i.v. daily for 2 days) and then underwent allogeneic stem cell transplantation. Busulfan clearance and area under concentration time-curve (AUC) were determined after the first dose using a one-compartment pharmacokinetic (PK) model with first-order absorption. The initial PK analysis was successfully completed after the first BU dose in 21 patients (84%). A final AUC of 1000-1500 microM x min/dose was targeted and subsequent doses were modified as necessary to achieve this value. Fourteen of the 25 patients (56%) required dose adjustment. Follow-up PK analysis was completed in 21 patients and 16 of these achieved the targeted BU exposure for the course of therapy. Interpatient variability in BU clearance was high (up to five-fold). The most frequent regimen-related toxicities were cutaneous and gastrointestinal (stomatitis and diarrhea). Only one patient developed hepatic veno-occlusive disease. Our study demonstrates the feasibility of adjusting the oral BU dose in individual pediatric patients. Although toxicity associated with BU seemed to be reduced, this conclusion is tempered by the fact that the overall regimen-related toxicity (RRT) remains substantial and reflected the effects of all agents used in the preparative regimen.