The availability of an i.v. form of busulfan (Bu) has prompted investigation of administration schedules other than the 4-times-daily dosage commonly used with oral Bu. We have studied an allogeneic stem cell transplantation (SCT) preparative regimen comprising fludarabine (FLU) 50 mg/m2 on days -6 to -2 plus i.v. Bu 3.2 mg/kg daily in a 3-hour infusion on days -5 to -2. The regimen was given to 70 patients aged 15 to 64 years (median, 41 years) with hematologic malignancy. Thirty-six patients (51%) had high-risk malignancy, 28 (40%) had unrelated or genotypically mismatched related donors (alternate donors [AD]) and 29 (41%) received bone marrow rather than blood as stem cell source. Acute GVHD prevention comprised antithymocyte globulin 4.5 mg/kg over 3 days pretransplantation, cyclosporin A, and short-course methotrexate with folinic acid. Hepatic toxicity was transient and there was no clinically diagnosed veno-occlusive disease. Grade II stomatitis occurred in 49 patients (70%) and hemorrhagic cystitis in 9 patients (13%). One patient with subtherapeutic phenytoin levels had a convulsion 8 hours after the third i.v. Bu dose, but no other neurotoxicity was apparent. Incidence of acute GVHD grades II to IV was 8% and incidence of grade III-IV was 3%, with no deaths from this cause. Actuarial incidence of chronic GVHD at 2 years is 38%. There were 2 cases of graft failure in unrelated donor BMT recipients, 1 of which was reversed by asecond transplantation. With a median follow-up of 16 months (range, 6-27 months), transplantation-related mortality at 100 days and 2 years was 2% and 5% for matched related donor (MRD) SCT and 8% and 19% for AD SCT, respectively (P = not significant). Relapse rates were 21% for 34 patients with acute myeloid leukemia (AML) in complete remission or chronic myeloid leukemia in chronic phase (low-risk), 66% for 19 patients with high-risk AML, and 18% for 17 patients with other active malignancy. Projected disease-free and overall survival rates at 2 years were 74% and 88% for low-risk disease, 26% and 37% for advanced AML, and 65% and 71% for other high-risk disease, respectively. Pharmacokinetic studies were done using 11 samples with the first and fourth doses of Bu. Kinetics were linear, and for the first and fourth doses, the half-lives were 2.60 +/- 0.44 and 2.57 +/- 0.36 hours, respectively. Clearances were 106.77 +/- 16.68 and 106.86 +/- 21.57 mL/min per m2, peak concentrations (Cmax) were 3.92 +/- 0.31 and 3.96 +/- 0.28 mcg/mL, and Bu areas under the plasma concentration versus time curve (AUC) were 4866.51 +/- 771.42 and 4980 +/- 882.80 microM x min, respectively. Bu was completely cleared within 24 hours and the day 4 pharmacokinetic values were very similar to those on day 1 for every patient. The cumulative AUC was comparable to the target range established for p.o. Bu. This regimen incorporating once-daily i.v. Bu is convenient to give, is relatively well tolerated, gives predictable blood levels, and deserves further study in circumstances in which cytoreduction as well as immune suppression is needed.