Abstract The aims of this study were twofold: (1) to assess the marrow of patients with T-lineage acute lymphoblastic leukemia (T-ALL) for the presence of molecular residual disease (MRD) at different times after diagnosis and determine its value as a prognostic indicator; and (2) to compare the sensitivity, rapidity, and reliability of two methods for routine clinical detection of rearranged T-cell receptor (TCR). Marrow aspirates from 23 patients with T-ALL diagnosed consecutively from 1982 to 1994 at the Division of Pediatric Hematology and Oncology, University of Catania, Italy, were obtained at diagnosis, at the end of induction therapy (6 to 7 weeks after diagnosis), at consolidation and/or reinforced reinduction (12 to 15 weeks after diagnosis), at the beginning of maintenance therapy (34 to 40 weeks after diagnosis), and at the end of therapy (96 to 104 weeks after diagnosis). DNA from the patients' marrow was screened using the polymerase chain reaction (PCR) for the four most common TCR δ rearrangements in T-ALL (Vδ1Jδ1, Vδ2Jδ1, Vδ3Jδ1, and Dδ2Jδ1) and, when negative, further tested for the presence of other possible TCR δ and TCR γ rearrangements. After identification of junctional rearrangements involving V, D, and J segments by DNA sequencing, clone-specific oligonucleotide probes 5′ end-labeled either with fluorescein or with [γ-32P]ATP were used for heminested PCR or dot hybridization of PCR products of marrows from patients in clinical remission. For 17 patients with samples that were informative at the molecular level, the estimated relapse-free survival (RFS) at 5 years was 48.6% (±12%). The sensitivity and specificity for detection of MRD relating to the outcome were 100% and 88.9% for the heminested fluorescence PCR and 71.4% and 88.9% for Southern/dot blot hybridization, respectively. Predictive negative and positive values were 100% and 90.7% for heminested fluorescence PCR, respectively. The probability of RFS based on evidence of MRD as detected by heminested fluorescence PCR at the time of initiation of maintenance therapy was 100% and 0% for MRD-negative and MRD-positive patients, respectively. Thus, the presence of MRD at the beginning of maintenance therapy is a strong predictor of poor outcome, and the molecular detection of MRD at that time might represent the basis for a therapeutic decision about such patients. By contrast, the absence of MRD at any time after initiation of treatment strongly correlates with a favorable outcome. The heminested fluorescence PCR appears to be more accurate and more rapid than other previously used methods for the detection of residual leukemia.