Relapse remains the major cause of treatment failure after hematopoietic cell transplantation (HCT) in acute leukemia, even in patients transplanted in morphologic CR. Various techniques now enable the sensitive quantification of 'minimal' amounts of residual disease (MRD) in patients with acute leukemia in remission. Numerous studies convincingly demonstrate that MRD at the time of transplantation is a powerful, independent predictor of subsequent relapse, with current detection levels of one leukemic cell in 10(5)-10(6) normal cells being prognostically relevant. This recognition provides the rationale to assign patients with detectable MRD (that is, 'MRD(+)' patients) to intensified therapies before, during, or after transplantation, although data supporting these strategies are still sparse. Limited evidence from observational studies suggests that outcomes with autologous HCT are so poor that MRD(+) patients should preferentially be assigned to allogeneic HCT, which can cure a subgroup of these patients, particularly if unmanipulated (T-cell replete) grafts and/or minimized immunosuppression are used to optimize the graft-vs-leukemia effect. Emerging data suggest that additional therapy with non-cross-resistant agents to decrease residual tumor burden before transplantation in MRD(+) patients might be beneficial. Further, other studies hint at immunotherapy (for example, rapid withdrawal of immunosuppression and/or donor lymphocyte infusions) as a means to prevent overt relapse if patients remain, or become, MRD(+) after HCT. Ultimately, controlled clinical studies are needed to define the value of MRD-directed therapies, and patients should be encouraged to enter such trials.