Despite aggressive therapy for leukemia in the form of bone marrow transplantation (BMT) relapse occurs in a significant number of cases. The origin of the leukemic relapse, whether it is of donor or recipient origin, and how best to treat the patients continue to pose problems for the clinician. In this paper we present a case in which the cytogenetics suggested that the relapse was of donor origin; however, molecular analysis revealed that the leukemic population was of host origin. The leukemic relapse following the BMT was treated with a second BMT. This resulted in a remission of 28 months after which leukemic relapse was again diagnosed. Using conventional chemotherapy it was possible to obtain another complete remission. This case illustrates a pitfall to cytogenetic analysis and two contrasting methods of dealing with leukemic relapse following BMT.