Considerable experimental data suggest that most, if not all, recipients of conventional (non-T cell-depleted) HLA-identical sibling transplants are likely to develop graft-versus-host disease (GVHD). This is because all donor-recipient pairs are likely to be disparate for numerous non-HLA antigens and because these grafts contain substantial numbers of donor T cells. However, about one-third to one-half of recipients of this type of transplant have little or no GVHD. Here we review recent data in humans suggesting that others factors, like the ability of the host to modulate GVHD reactivity of immune competent cells in the graft, may explain lack of GVHD in some persons. This concept may have implications for preventing and treating GVHD in humans.