Previous work from this laboratory demonstrated that tolerance to MHC-identical or class I-disparate renal allografts develops in approximately one third of miniature swine without exogenous immunosuppression. A back-cross study indicated that rejection of MHC-identical transplants due to minor Ag was controlled by one or possibly two non-MHC-linked, autosomal dominant Ir genes. According to this hypothesis, and assuming complete penetrance, graft acceptors would be homozygous recessive at the relevant Ir loci, as would their offspring. Alternatively, if the gene(s) were incompletely penetrant, then two acceptors could give rise to a rejector. However, a high rate of MHC-identical graft acceptance would still be expected in the offspring of acceptors even if the Ir gene(s) were incompletely penetrant. To test this hypothesis and to obtain a higher frequency of acceptor animals for studies of tolerance, a program of selective breeding of renal allograft acceptors was begun. In the present paper, we assess the effect of selective breeding on renal graft acceptance. The analysis indicates a marked increase in the rate of MHC-identical graft acceptance, from 27.3% (n = 24) for the earliest of the four chronologic subgroups assessed to 64.5% (n = 33) for the most recent subgroup (p less than 0.0001). Calculations of kinship revealed that the increased acceptance of MHC-identical grafts was not the result of differences between acceptors and rejectors in donor/recipient consanguinity. Class I-disparate grafts (n = 128) were similarly stratified chronologically and compared. Unlike MHC-identical grafts, the rate of acceptance of class I-disparate grafts has not changed over time. We conclude that rejector/acceptor status with respect to class I MHC incompatibility is determined by genetic factors in addition to those that control responses to minor antigen incompatibilities only.