We evaluated the efficacy of a treatment strategy in which infants with acute lymphoblastic leukemia (ALL) were stratified by their MLL gene status and then assigned to different risk-based therapies. A total of 102 patients were registered on two consecutive multicenter trials, designated MLL96 and MLL98, between 1995 and 2001. Those with a rearranged MLL gene (MLL-R, n=80) were assigned to receive intensive chemotherapy followed by hematopoietic stem cell transplantation (HSCT), while those with germline MLL (MLL-G, n=22) were treated with chemotherapy alone. The 5-year event-free survival (EFS) rate for all 102 infants was 50.9% (95% confidence interval, 41.0-60.8%). The most prominent late effect was growth impairment, observed in 58.9% of all evaluable patients in the MLL-R group. This plan of risk-based therapy appears to have improved the overall prognosis for infants with ALL, compared with previously reported results. However, over half the events in patients with MLL rearrangement occurred before the instigation of HSCT, and that HSCT-related toxic events comprised 36.3% (8/22) of post-transplantation events, suggesting that further stratification within the MLL-R group and the development of more effective early-phase intensification chemotherapy will be needed before the full potential of this strategy is realized.