Previous studies from our laboratory have already shown that pretransplant blood transfusions from third-party donors significantly prolong kidney allograft survival in mismatched unrelated immunosuppressed dogs. In the present study, a similar effect was found in related donor-recipient pairs mismatched for two haplotypes (P less than 0.05); no significant effect was observed in one-haplotype-mismatched combinations (P = 0.34). Pretransplant blood transfusions did not have a beneficial influence on kidney allograft survival in immunosuppressed DLA-identical, related donor recipient pairs. After withdrawal of immunosuppressive therapy, transfused dogs in this group rejected their kidneys even more frequently than did nontransfused dogs (P = 0.16). A comparable undesired effect of blood transfusions was found for recipients of unrelated kidneys which were identical with respect to DLA-A, B, and D: transfused recipients of those kidneys rejected the graft significantly more often than the untransfused controls (P less than 0.01). The most likely explanation for this adverse effect is that blood transfusions given to the recipient may cause crossimmunization for undefined, probably minor, antigens of the donor kidney. Apparently, differences arising from these minor histocompatibility antigens become manifest only after withdrawal of the immunosuppressive therapy. Furthermore, it appeared that the effect of histocompatibility matching on kidney allograft survival is less for transfused than for untransfused dogs. The most important conclusion is, however, that the beneficial effect of blood transfusions appears to be dependent on the degree of matching: while blood transfusions are on the whole beneficial for unmatched kidneys, they are likely to have no effect, or even to be harmful, for matched kidneys.