The transplantation of organs between individuals of a species normally has two main consequences: (1) the tissue is rejected unless the individuals are matched for identity of transplantation antigens, especially those encoded by the major histocompatibility complex (MHC), and (2) the recipient is sensitized to the transplantation antigens of the donor so that second-set grafts are rejected in a more violent manner¹. These rules of transplantation are not obeyed by liver grafts. Orthotopic transplants of liver are never rejected by many strains of rat even though the MHC barrier is crossed². We have recently shown that instead of sensitizing the recipient, these enduring liver grafts induce a state of donor-specific unresponsiveness in which subsequent grafts of other organs, such as skin, are accepted permanently. It is possible that many strains simply cannot mount a sufficiently vigorous destructive immune response to outstrip the liver's great capacity to repair immune damage and that the systemic unresponsiveness observed is due to diversion of alloreactive lymphocytes with donor specificity into the liver allograft. Manipulations which increase the vigour of the immune response might therefore tip the balance in favour of liver graft rejection and away from induction of unresponsiveness. We have previously measured the degree to which the immune response can be increased by previous exposure to MHC antigens in a quantitative adoptive transfer system³. In the strain combination DA (MHC haplotype RT-1^a) grafted to PVG (RT-1^c), the lymphocytes of immunized animals are 1,000 times more potent than those of non-immune animals in procuring graft destruction when transferred to irradiated hosts. Unexpectedly, we have now found that liver grafts transplanted into previously immunized rats not only fail to reject but convert the state of heightened reactivity to donor grafts characteristic of immune recipients into one of non-reactivity characteristic of tolerant animals.