In humans, the success rate of bone marrow transplantation (BMT) across major histocompatibility complex (MHC) barriers is not high due to: (1) graft-versus-host reaction (GvHR); (2) graft rejection, and (3) incomplete T cell recovery. In mice, GvHR can be prevented if T cell- depleted bone marrow cells (BMCs; <2% T cells) are used. Graft rejection can be prevented by either bone grafts (to recruit donor-derived stromal cells) or the injection of donor BMCs via the portal vein (p.v; to induce donor-specific tolerance). T cell functions are recovered by BMT plus bone grafts if the thymic functions of recipients are not completely lost. After the complete loss of thymic functions (due to aging), BMT plus embryonal thymus grafts should be carried out.Recently, we have found that persistent donor-specific tolerance can be induced if allogeneic hemopoietic stem cells are injected via the p.v. Based on these findings, we have established new strategies for organ allografts. Without irradiation, donor BMCs should be injected from the p.v. injection on day 0 plus i.v. injection on day 5, and an immunosuppressant (CsA or FK506) should be used on days 2 and 5. Without using immunosuppressants, sublethal irradiation (7 Gy) followed by skin allografts plus allogeneic BMC injection via the p.v. should be carried out. This leads to a 100% acceptance of skin and pancreas allografts for more than 300 days. The recipient mice show mixed allogeneic chimerism, and spleen cells from the recipients show tolerance to both donor-type and host-type MHC determinants in the assays for mixed lymphocyte reaction and generation of cytotoxic T lymphocytes. We have confirmed that these strategies are applicable to other animals such as pigs and rats. We therefore believe that they will become viable and valuable strategies for human organ allografts.