Pretreatment of prospective donors of hemopoietic cells with a single recipient-specific blood transfusion can significantly decrease the morbidity and mortality of graft-vs.-host disease (GVHD) in lethally irradiated, allogeneically reconstituted mice. This beneficial effect of donor pretreatment could be demonstrated in donor-recipient strain combinations that were H-2 + non-H-2, H-2, or only class II-disparate, but not in the class I-disparate C57BL-B6.C-H-2bm1 strain combination. The effect was proportional to the amount of recipient-strain blood used for transfusion. Donor transfusion with a single dose of 1 ml recipient-specific whole blood resulted in minimum GVHD, lower doses being less or not effective. The interval between donor pretreatment and the use of their hemopoietic cells for reconstitution appeared to be important. The best survival was found at an interval of 4 days. Multiple transfusion was not more effective than a single one. We compared the effectiveness of whole blood and irradiated spleen cells for donor pretreatment. Both protocols have been shown previously to suppress anti-recipient delayed-type hypersensitivity. It appeared that the blood transfusion protocol was superior to the spleen cell protocol. The beneficial effect appeared to be recipient specific, since a third-party blood transfusion did not improve GVHD. We found that the beneficial effect of donor blood transfusion was due to suppression of the anti-host immune response. The donor blood transfusion was able to induce bystander suppression to alloantigens that were not used for the induction of suppression, provided they were co-expressed with the specific alloantigens by the recipients. This also indicates that, although the induction of suppression is specific, the ultimate suppressive effect is non-specific.