BACKGROUND: The initial response to induction therapy is currently considered one of the most important prognostic factors in acute lymphoblastic leukemias (ALL). A series of methods for the detection of submicroscopic levels of residual disease in patients with ALL mainly based on PCR and immunophenotyping has been developed, demonstrating that the presence of high levels of residual disease at the end of induction therapy is an important, independent prognostic factor. We determined the usefulness of PCR detection of minimal residual disease using consensus primers as a non-remission criterion. PROCEDURE: Bone marrow samples obtained from 49 children with ALL were analyzed at diagnosis and at the end of induction therapy for the detection of clonal IgH, TCRdelta, and TCRgamma rearrangements by PCR. The results were compared with those obtained by standard morphologic analysis and risk group classification. RESULTS: Patients who had clonality detected at the end of induction showed a significantly higher recurrence rate and lower event-free survival than those without detected clonality (24.9% vs. 89.7%) (P < 0.0001). Multivariate analysis revealed that detection of clonality at the end of induction was the most important, independent prognostic factor when associated with age, number of white blood cells, and immunophenotyping. CONCLUSIONS: PCR detection of clonality using consensus primers is a relatively simple technique that is able to identify patients with a high chance of recurrence, and shows a higher sensitivity and a better prognostic value than standard morphologic analysis and risk group classification, defining a new remission criterion. However, further multicentric prospective studies using this technique employing a larger number of cases are necessary to confirm these findings.