Prevention of rejection and the induction of transplantation tolerance are two related but separable phenomena that must both be considered in the analysis of the response to a transplanted organ. It is frequently hard to separate these phenomena in assessing the outcome of clinical transplants, because patients are rarely studied in the absence of immunosuppressive agents. Use of our partially inbred miniature swine has permitted us to examine the effects of selective MHC matching on transplant survival, and the data indicate that matching has an effect on both phenomena. Prevention of early rejection with CyA was possible for all mismatches examined, although it was clearly more difficult with increasing degrees of mismatching. On the other hand, tolerance induction after cessation of the immunosuppressive agent was dependent on presence of at least one matched MHC locus between the donor and recipient, with complete class II matching appearing to be the most successful way of assuring long-term graft survival. It is also apparent from our data that although durable tolerance to primarily vascularized renal allografts could be induced across a variety of selective MHC disparities, all cases involving a class II mismatch (ie, selective class I matched or one-haplotype full MHC mismatched kidney allografts) underwent spontaneously reversible rejection crises during the early follow-up period. Such a clinical course might be unacceptable for human clinical trials, even though the transient renal dysfunction may reflect events involved in tolerance induction rather than true rejection (Gianello et al: Immunol Rev 133:19, 1993.). Indeed, we do not yet know whether or not further immunosuppressive treatment at the times of such crises may prevent rather than facilitate the induction of tolerance. On the other hand, in the case of selective two-haplotype class I mismatch the regimen utilized was capable of inducing tolerance to renal allografts in 100% of the recipients with minimal or no renal dysfunction throughout the follow-up period. Although the excellent results achieved with current antirejection agents has led to debate about the wisdom of HLA matching for cadaver transplants in terms of preventing rejection, our data would suggest that such matching might be of even greater importance for success of protocols in which attempts are made to induce transplantation tolerance. Because class II antigens are less polymorphic than are class I antigens, mismatching for class I antigens may be achievable for cadaver donor transplantation, and may provide the first situation in which these principles can be applied to clinical trials.