A chimeric severe combined immunodeficient mouse engrafted with human peripheral blood (hu-PBL-SCID) model has been developed to test anti-T-cell monoclonal antibody (mAb) effects on systemic symptoms of the host and the survival of human skin grafts. To obtain consistent engraftment without lethal acute graft-versus-host disease (GVHD), SCID mice were pretreated with a combination of total body irradiation (2.5 Gy, day 0) and anti-asialo GM1 (anti-mouse natural killer cell) antiserum (50 micrograms i.p., day 3) before the intraperitoneal injection of 40-50 X 10(6) human PBL on day 4. With this protocol, the engraftment rate was 82% with 5-98% human CD45-positive cells in the peripheral blood. Mortality at 30 days was 0% in the mice bearing 5-50% human cells compared with 70% in those with more than 50%. Using hu-PBL-SCID mice with 5-50% human cells in their peripheral blood, we demonstrated the following results: 1) Human T cells isolated from these mice proliferated in response to immobilized OKT3 stimulation in vitro. 2) Hu-PBL-SCID mice but not normal SCID mice were able to reject human skin grafts in vivo 16-21 days after grafting. 3) Both OKT3 (anti-human CD3 mAb) and T10B9 (anti-human alpha beta T-cell receptor mAb) treatment prevented human skin graft rejection in hu-PBL-SCID mice. 4) OKT3 but not T10B9 induced first dose reactions characterized by hypothermia and hypoactivity which were consistently observed within 90 min of intravenous injection into hu-PBL-SCID mice. 5) Human cytokines were detected in the serum of the hu-PBL-SCID mice treated with anti-T-cell mAbs. The close similarity of these responses to human clinical mAb immunosuppressive therapy suggests that the hu-PBL-SCID mouse model may be an excellent tool for investigating the immunosuppression, side effects, and mechanism of action of agents that are specific for human and higher apes and not reactive with lower animals.