Alterations in hemostasis leading to symptomatic thromboembolism have been observed in patients with acute lymphoblastic leukemia (ALL) receiving Escherichia coli asparaginase (CASP) combined with steroids. Moreover, hereditary prothrombotic risk factors are associated with an increased risk for venous thromboembolism in pediatric ALL patients treated according to the BFM 90/95 protocols (including CASP combined with prednisone during induction therapy). To assess whether the thromboembolic risk associated with established prothrombotic risk factors is modified by treatment modalities (prednisone or dexamethasone), the present analysis was performed. Three hundred thirty-six consecutively recruited leukemic children treated according to different BFM protocols (PRED group, n ؍ 280, 60 mg/m 2 prednisone; DEXA group, n ؍ 56, 10 mg/m 2 dexamethasone during induction therapy) were studied. Study end point was the onset of symptomatic vascular accidents during induction therapy. Cumulative thromboembolism-free survival was significantly reduced in children in the PRED group (thrombosis frequency, 10.4%) compared with children in the DEXA group (thrombosis frequency, 1.8%; P ؍ .028). Although no significant difference was found in the overall prevalence of prothrombotic risk factors, 46.5% of patients in the PRED group who experienced thromboembolic events were carriers of a prothrombotic risk factor, whereas no carrier in the DEXA group had a thromboembolism. At the time of maximum CASP activity, fibrinogen and activities of antithrombin, plasminogen, and protein S were significantly reduced in the PRED group. No significant correlation could be found between CASP activity and levels of coagulation factors. In conclusion, the use of dexamethasone instead of prednisone, administered with CASP, significantly reduced the onset of venous thromboembolism. (Blood. 2003;101:2529-2533)