Recipient dogs were conditioned with 9.2 Gy of total-body irradiation followed by the infusion of bone marrow and peripheral blood leukocytes from a DLA-haploidentical litter mate (N=1O) or a completely allogeneic unrelated donor (n=9). Graft-vs.-host disease (GVHD) prophylaxis consisted of methotrexate (MTX) and cyclosporine (CsA). Postgrafting all dogs were complete lymphohemopoietic chimeras. Lymphocytes of haploidentical chimeras without GVHD were unresponsive to stimulation by host lymphocytes cryopreserved pregrafting. Lymphocytes of haploidentical chimeras with GVHD proliferated in response to host cells, albeit less than donor cells pregrafting. In completely allogeneic chimeras, neither lymphocytes from dogs with GVHD, nor those from dogs without the disease showed responses to host lymphocytes. In addition, cells from haploidentical chimeras obtained early after transplantation non-specifically suppressed donor cell proliferation. Later on, lymphocytes from dogs without GVHD showed specific suppression of donor cells, while lymphocytes from chimeras with GVHD continued to show nonspecific suppression. Cells from completely allogeneic chimeras both with and without GVHD never suppressed donor cells specifically. Both specific and nonspecific suppressor cells were enriched by nylon wool adherence, expressed T cell markers, and were not affected by the addition of indomethacin. Even after removing nylon wool—adherent cells, however, chimera cells were unresponsive to stimulation by host cells. By one year after transplant, chimera lymphocytes no longer showed suppression. In cell-mediated lympholysis assays, lymphocytes from all chimeras, regardless of GVHD, failed to generate cytotoxic cells against host cell targets. However, while haploidentical chimeras showed cytotoxicity against third-party targets, completely allogeneic chimeras did not. This deficiency was not overcome by the addition of mixed leukocyte culture supernatant or donor lymphocytes. All chimeras had basically normal antibody responses to keyhole limpet hemocyanin and phage X174. However, while haploidentical chimeras had normal responses to bacillus Calnette-Guerin (BCG) sensitization and rejected DLA-incompatible skin grafts within the normal time frame, completely allogeneic chimeras were not sensitized by BCG and showed delayed skin graft rejection. Histopathological studies revealed slow thymic reconstitution in all chimeras, particularly in the presence of GVHD. However, while healthy haploidentical chimeras eventually showed thymic histology normal for age, completely allogeneic chimeras did not. In agreement with previous studies, these data are compatible with the notion that clonal abortion is involved in the development of tolerance. In addition, different patterns of alloreactivity are associated with the presence of specific or nonspecific suppressor cells. While haploidentical chimeras become immunocompetent, major defects of cellular immunity persist in completely allogeneic chimeras.