Disease relapse after allogeneic bone marrow transplantation (BMT) is a major cause of treatment failure and is thought to evolve from clinically occult residual disease in the recipient. However, the demonstration of minimal residual disease (MRD) in individual patients is of uncertain prognostic significance because the detection of residual disease has not consistently correlated with subsequent relapse. Moreover, the optimal therapeutic approach in patients with MRD after allogeneic BMT is unknown. The study of these issues has been hindered by the lack of clinically relevant animal models. In this report, we characterize a novel murine model for the study of MRD after allogeneic BMT. This model was designed to simulate high-risk BMT in humans in which patients receive transplants in relapse and disease recurrence in the major cause of treatment failure. The H-2-compatible, mixed lymphocyte culture nonreactive murine strains, AKR (H-2k) and CBA (H-2k), were chosen to parallel marrow transplants from HLA-matched siblings, which represent the majority of allo-transplants in humans. Male AKR leukemia cells were used in female donor/host chimeras permitting the Y chromosome to serve as a leukemia-specific marker for MRD. Detection of residual male leukemia cells in the peripheral blood of the primary host was facilitated by use of the polymerase chain reaction (PCR) and sequence-specific oligonucleotide probe hybridization (SSOPH). Use of PCR/SSOPH was highly predictive of clinical outcome (relapse or cure) in animals receiving transplants (P < .00002) and detected disease recurrence earlier than comparative flow cytometric analysis studies. This murine model will be useful in evaluating the efficacy of therapeutic strategies aimed at reducing disease relapse posttransplant and can be adapted to other transplant murine tumor systems for the study of MRD.