Tumor relapse remains a major obstacle to the success of allogeneic bo
ne marrow transplantation (BMT) as a treatment for leukemia. Due to li
mited treatment options, the outlook for most patients that relapse fo
llowing allogeneic BMT has been poor. The infusion of normal immunocom
petent leukocytes from the original marrow donor has become a promisin
g new option for treating/preventing leukemia relapse in allogeneic BM
T recipients. This form of treatment has often been referred to as don
or leukocyte infusion (DLI) therapy. Our laboratory is using murine mo
dels of allogeneic BMT to address important unresolved issues regardin
g DLI therapy in an effort to make the treatment more effective. These
include identification of the antileukemic effector cells, aug mentat
ion of the antileukemic effect, and understanding why graft-versus-hos
t-disease (GVHD) is less severe than anticipated. This article reviews
our work in murine models of DLI and introduces our current working h
ypotheses concerning DLI therapy.