We have previously shown that blood transfusions can immunize a dog and lead to rejection of a subsequent marrow graft despite lethal total body irradiation (TBI). Sensitization to histocompatibility antigens induced by two prior transfusions of whole blood could be overcome by a regimen of procarbazine and anti-thymocyte serum (ATS) preceding TBI. The current study investigated a) whether this regimen could abrogate sensitization induced by six weekly transfusions given from days --50 to --15 preceding a marrow graft, and b) whether platelet survival studies and two in vitro tests of immunity could predict marrow graft rejection. All donor-recipient pairs were histoincompatible, unrelated, and of different breed. Twenty-two recipients received platelet concentrate transfusions and eight received whole blood transfusions. Recipients were given 1200 R TBI and a graft of marrow and peripheral blood leukocytes from the transfusion donor on day 0. Three of 15 recipients (20%) given procarbazine, 12.5 mg/kg i.v. on days --8, --6, and --4, and ATS, 0.6 ml/kg subcutaneously on days --7, --5 and --3, Rejected their grafts, whereas 11 of 15 dogs (73%) not given procarbazine and ATS rejected their grafts (p less than 0.01). Serum lymphocytotoxic antibodies, peripheral leukocyte migration inhibition, and in vivo donor platelet recovery and survival were studied in those recipients receiving six weekly transfusions and in 18 other recipients receiving a single donor transfusion 3 months before marrow grafting. No significant correlation was found among these in vitro and in vivo tests of sensitization. Sensitization to marrow grafts was not reliably detected by the presence of cytotoxic antibodies or leukocyte migration inhibition. Platelet survival, however, was positively correlated with the results of marrow grafting in 12 of 15 (80%) evaluable recipients (p approximately 0.15).