Intravenous busulfan (i.v. BU) has demonstrated safety when administered at 0.8 mg/kg per dose i.v. every 6 hours x 16 doses. We evaluated the safety and pharmacokinetics (PK) of giving the same total daily i.v. BU dose (3.2 mg/kg) either divided as a twice-daily infusion or as a single infusion to patients undergoing hematopoietic stem cell transplantation (HSCT). Twelve patients with hematologic malignant disease were treated; 7 patients had non-Hodgkin's lymphoma, 4 patients had acute myeloid leukemia, and 1 patient had chronic myelogenous leukemia. The first cohort (group A) received, on the basis of actual body weight, i.v. BU at 1.6 mg/kg per dose over 4 hours every 12 hours for 4 days (day -7 to day -4). The second cohort (group B) received 3.2 mg/kg per dose of i.v. BU (same total dose as group A) as a single infusion over 4 hours daily for 4 days. In both groups the i.v. BU was followed by cyclophosphamide 60 mg/kg daily for 2 days (day -3 and day -2). Blood specimens were collected on the first, fifth, and seventh doses for group A and on the first and fourth doses for group B to determine the disposition of i.v. BU. Peripheral blood stem cells (autologous in 7 cases and HLA-matched allogeneic in 5 cases) were given 2 days after completion of cyclophosphamide administration (day 0), and granulocyte colony-stimulating factor 5 microg/kg was started on the same day. GVHD prophylaxis consisted of tacrolimus plus methotrexate for recipients of allogeneic stem cells. One patient developed presumed fungal pneumonia and died of multisystem organ dysfunction on day +21 before hematologic reconstitution could be evaluated. Another was reported to have sudden death of undetermined cause at home on day 40. The remaining patients had engraftment (absolute neutrophil count >500/microL) at a median of 11 days and sustained platelet counts >20,000/microL at a median of 14 days. Significant regimen-related toxicity (grade III-IV) was limited to hepatic toxicity (2 cases) catheter infection (2 cases), epistaxis (3 cases), diarrhea (1 case), anorexia (1 case), mucositis (1 case), hyperglycemia (1 case), pneumonia (1 case), and sepsis (1). In group B there was 1 case of mild venoocclusive disease, which resolved without sequelae. No central nervous system or pulmonary toxicity was noted. Pharmacokinetic parameters, including clearance, half-life, maximum concentration, and area under the curve, demonstrated that the first dose profile was highly predictive of later dose PK profiles. No accumulation of the drug was noted. The change in dosing schedule did not increase toxicity or end-organ damage despite higher plasma concentration-times. Although further study for long-term efficacy is warranted, i.v. BU can be given safely with reproducible results on a twice-daily divided or single-daily dosing schedule to patients undergoing HSCT.