The purpose of the study was to evaluate the effect of lisofylline (LSF) on engraftment, regimen-related toxicities (RRT), and mortality in patients undergoing allogeneic bone marrow transplantation (BMT). We performed a multicenter, randomized placebo-controlled trial in 60 patients with hematologic malignancies receiving BMT from HLA-identical sibling donors. Patients were randomized to receive either placebo, 2 mg/kg LSF or 3 mg/kg LSF every 6 h, beginning before conditioning and continuing to day 21 or hospital discharge. Treatment groups were balanced with respect to conditioning regimen and disease stage. However, significantly more patients in the 2 mg/kg LSF group were at high risk for RRT due to performance status >/=1, age >/=40 years, and prior exposure to CMV. Nausea and vomiting were the only adverse events observed in a higher proportion of LSF-treated patients that led to study withdrawal in six of 42 patients (14%). The times to neutrophil recovery to >/=500/microl and platelet recovery (>20 000/microl) were not improved by LSF treatment. Nevertheless, no patient who received treatment with 3 mg/kg LSF developed a documented infection between day 0 and 35 or had a serious or fatal infection between day 0 and 100 (P = 0.003 vs placebo for both). The day-100 survival rate was also significantly improved in the 3 mg/kg LSF group (89%), compared with either the 2 mg/kg LSF (48%) or placebo (61%) groups (log-rank test, 3 mg/kg LSF vs placebo, P = 0. 026). We conclude that treatment with LSF 3 mg/kg reduced the incidence of infections and improved 100-day survival in patients receiving related-donor allogeneic bone marrow transplantation. Bone Marrow Transplantation (2000) 25, 283-291.