Conventional preparative regimens for allogeneic stem cell transplantation are associated with excessive regimen-related toxicity (RRT) in some patients because of underlying comorbidities, advanced age, or prior treatment. We studied a preparative regimen designed to reduce RRT, yet allow for adequate engraftment and development of a graft-versus-malignancy effect. Thirty patients (median age, 57 years) were entered on study. Twenty-nine patientsreceived stem cells from HLA-identical siblings and 1 from a sibling mismatched for 1 antigen at the A locus. Sixteen patients had received previous stem cell transplants (6 allogeneic and 10 autologous). The preparative regimen consisted of fludarabine 30 mg/M2 per day IV on day -10 to day -5, busulfan 1 mg/kg per dose PO (n = 6) or 0.8 mg/kg per dose IV (n = 24) for 8 doses every 6 hours on day -6 to day -5, and horse-derived antithymocyte globulin 5 mg/kg per day IV (n = 12) or 15 mg/kg per day IV (n = 18) on day -4 to day -1. GVHD prophylaxis consisted of cyclosporine (CYA) 3 mg/kg BID PO starting on day -3 (n = 13) or CYA and methotrexate 15 mg/m2 IV on day +1 and 10 mg/m2 IV on day +3 and day +6 (n = 17). The median number of CD34 cells transplanted was 3.19 x 10(6)/kg. All patients demonstrated recovery of hematopoietic function. Twenty-six (89%) of 29 evaluable patients achieved greater than 90% donor cell chimerism before day 100. Three patients never achieved greater than 90% donor chimerism, and another 3 patients subsequently lost donor chimerism. All 6 of these patients had autologous reconstitution with progressive disease. RRT was minimal; 7 patients had greater than grade II nonhematologic toxicity and there were no toxic deaths attributable to the conditioning regimen. Transplantation-related mortality was 7% (95% confidence interval [CI], 6%-8%) at 3 months and 28% (95% CI, 23%-34%) at 12 months after transplantation. Non-relapse-related mortality was most often due to infection. Grade II or greater GVHD developed in 56% of evaluable patients, and all patients with disease response developed GVHD. Actuarial estimates of overall and disease-free survival at 12 months were 52% (95% CI, 43%-63%) and 30% (95% CI, 24%-37%), respectively. Although this preparative regimen allowed adequate engraftment with minimal RRT, GVHD and infectious complications caused significant morbidity and mortality. Further study to define appropriate patient populations for this regimen, while limiting GVHD and infection risks, is needed.