405 children with acute lymphoblastic leukemia were stratified according to age, initial leucocytes count, lymph nodes, liver and spleen size, into three prognostic classes I, II, III. Protocol 08 LA 74 which they were applied included: 1)initial randomization between Prednisone, Vincristine, Daunorubicin or the same plus Cyclophosphamide for induction and reinductions; 2)doses adjustments to prognostic factors, increased doses being given to increased risk patients; 3)comparison between intrathecal Methotrexate and intrathecal Methotrexate plus Ara-C in addition to skull irradiation for CNS prophylaxis; 4)L-Asparaginase consolidation for all patients; 5)maintenance by 6-Mercaptopurine and Methotrexate in all patients and reinductions. The most striking conclusions to date are the improvement for increased risk patients, the frequency of primary testicular relapses contrasting with the low rate of meningitis, the prognostic implication of sex, the influence on remission duration of the number of courses necessary to achieve complete remission, the importance of using Cox Method to improve the identification of prognostic groups.