To improve the outcome of children with acute lymphoblastic leukemia (ALL) treated at the National Cancer Institute, Bogota, Colombia, a protocol based on the BFM-90 (Berlin, Frankfurt, Munster study) and the LSA2L2 regimens was implemented in the year 1993. The patients were classified as being standard risk (SR) or high risk (HR) according to clinical criteria, to which cytogenetic information and day-8 prednisone response were also added. A 123-patient cohort entered the study, 18 of them being considered SR and 105 HR. There was a 94% 10 years' event-free-survival rate for the SR group and 36% for the HR group. Decreased induction death rate (7% vs. 14%), increased complete remission (CR) rate (81% vs. 75%), and continuous CR (45% vs. 33%) were found in comparison with the previous study. A significant improvement was achieved in relapse rate, 44% to 28% (P=0.029), mainly due to reduced central nervous system relapse rate from 16% to 6% (P=0.037), whereas the number of patients receiving cranial radiation was reduced to 55%. A major problem concerned the increased CR mortality rate, 5% to 14% (P=0.06). Improved supportive care therapy and socioeconomic conditions will hopefully reduce the CR mortality rate in the future.