The study of minimal residual disease has been fueled by the technologic advent of the polymerase chain reaction and basic developments identifying the genetic lesions involved in human malignancies. Thus far advances in identifying, cloning, and the subsequent polymerase chain reaction amplification of relevant genes have outpaced clinical studies designed to tell us the significance of minimal residual disease. It has become clear that the mere detection of minimal residual disease does not foretell relapse; thus, although the presence of minimal residual disease in acute lymphoblastic leukemia and acute prolymphocytic leukemia appears to be associated with a high risk of relapse, the presence of minimal residual disease in t(8;21) acute myeloid leukemia, chronic myeloid leukemia, and t(14;18) non-Hodgkin's lymphoma is not clearly associated with impending relapse. In most situations there is clearly a need for carefully controlled studies to evaluate the predictive value of minimal residual disease.