Up to 80% of pediatric patients with acute lymphoblastic leukemia (ALL) can be cured if intensive therapy is applied. Severe side effects are encountered in all patients of which, however, only the minority is life-threatening. The leading cause of failure in childhood ALL is still recurrence of disease. To reduce the rate of relapses, but also to limit treatment morbidity, the ALL-BFM group has aimed to improve the risk-adaptation of therapy. The most important addition to clinical factors (eg age, WBC, extramedullary involvement), and biological characteristics (such as immunphenotype and cytogenetics), was the recognition of early in vivo treatment response as the strongest predictor for relapse. The determination of leukemic blasts in peripheral blood after exposure to 7 days of prednisone (PRED) and one dose of intrathecal methotrexate (prednisone response) as developed by BFM identified multidrug resistant patients: Such patients had still more than 1,000 blasts per microL at day 8 of therapy (defined as PRED poor responders, 10% of all patients). Prognosis for these was only approximately 35% as compared to approximately 80% in patients with adequate PRED response. Patient characteristics at relapse reveal that most of them were originally comprised in" good risk" patient subgroups: eg, in trial ALL-BFM 90, 50% of the relapses were noted in patients with c-ALL even though that group had an EFS of 82%(SE 1%). 70% of the recurrences are found among patients with good response to PRED indicating the lack of specificity in the definition of that subgroup. Therefore, the more refined way of determining in vivo response based on the detection of minimal residual disease (MRD) at defined timepoints by identifying clone-specific T-cell receptor-(TCR) or immunglobuline (Ig) gene rearrangements appears to be able to define the patient at high risk to relapse more specifically. In the current ALL-BFM strategy, the high sensitivity of the method is utilized to apply treatment reduction in patients with fast clearance of leukemia. Persistent disease in contrast is an indication for treatment modification and intensification. Logistics and quality controls are demanding but essential for the introduction of this new technology into clinical practice.