For patients with standard-risk leukemia scheduled for an HLA-identical family donor transplant, T-cell depletion of the marrow graft constitutes an acceptable, if not superior, alternative to standard GvHD prophylaxis with cyclosporine/methotrexate, provided appropriate graft rejection prophylaxis with in vivo T-cell depletion or TLI is used. Although there is no evidence available to show that T-cell depletion is superior to conventional GvHD prophylaxis, in terms of leukemia-free survival, the more effective prevention of chronic GvHD results in a better quality-of-life for surviving patients. It is unclear whether replacing BM by blood as a stem-cell source improves the overall outcome in these good-risk patients. T-cell depletion of the graft is also appropriate for patients with high-risk features grafted in first or second CR, provided that an intensified conditioning regimen is used to reduce the risk of relapse. Patients with more advanced disease with a matched family donor or MUD should only receive a T-cell depleted graft in the context of a clinical trial geared to reducing the risk of relapse. In this setting, a PBPC graft may be preferable to a BM graft. If no compatible family donor is available and a compatible unrelated donor cannot be identified within an adequate time-frame, a T-cell depleted haploidentical PBPC graft is probably the most suitable approach, provided an intensified conditioning regimen and appropriate graft rejection prophylaxis are used. The delay in immune reconstitution remains the most intractable problem associated with T-cell depleted allogeneic SCT and a major research effort will be required to address this problem.