We examined the effect of high-dose methylprednisolone (HDMP) on blast reduction rate and compared it to conventional dose steroid treatment, administered during the first 7+ days of the induction remission period in patients with acute lymphoblastic leukemia (ALL). In our previous randomized study, the event free survival (EFS) was found to be higher in patients treated with HDMP (Group B) than in patients treated with a conventional dose of steroids (Group A). We used the chemotherapy protocol for ALL patients according to the St Jude Total XI study group. Medical records of all 194 patients who achieved complete remission were reviewed to determine the absolute blast count (ABC) in peripheral blood smears from day 0 of treatment until day 8. A target response of 1000 blast/mm 3 was employed. The time in which ABC remained greater than or equal to1000/mm 3 and the EFS rates were assessed in group A ( n =44) and B ( n =51) of high risk patients. Group A and B patients were branched to chemotherapy schedules (A1, B1:steroid+vincristine+daunorubicine+intrathecal+ 3 doses l -asparaginase; A2, B2: steroid+vincristine+daunorubicine+intrathecal; A3, B3: only steroid monotherapy). Number of patients from whom the ABC remained greater than or equal to1000/mm(3) during all 7 days of induction treatment and the median number of days for blasts to reach the target level were significantly lower in group B1 than A1 ( p <0.05), concordantly the 8-year EFS was higher in group B1 ( p =0.01). In comparison of the subgroups, results of A2 and A3 were worse than of group B2 and B3. These findings strongly suggest that the effects of HDMP on blast cytoreduction and EFS are more potent than conventional dose steroids.