A murine model of heterotopic corneal transplantation has been developed. Whole corneas were removed from either inbred Balb/cJ or C57B1/6J mice and inserted into an abdominal subcutaneous pouch in C57B1/6J recipient mice. Fresh donor tissue, donor corneas cultured at 37°C in 5 per cent CO2-95 per cent room air, or donor corneas cultured at 37°C in two atmospheres of 98 per cent 02- 2 per cent CO2(hyperbaric oxygen) were utilized. Twenty-one days later, the recipients were sacrificed and the heterotopic grafts were examined in a masked fashion. Rejected grafts were edematous, opaque, and vascularized. In contast, unrejected grafts were clear and nonvascularized. These results were further confirmed by histological examination of the donor grafts in a masked fashion. Control syngeneic grafts were not rejected. The fresh allografts were usually rejected (91%) as were the allografts cultured in CO2- room air (90%). Thus, this murine model of heterotopic corneal transplantation appears to be useful for the study of the factors involved in rejection.In further studies, the donor tissue was carried in hyperbaric oxygen in order to selectively destroy cells bearing la antigens (Langerhans cells and passenger leukocytes). These allografts were rejected only 23 per cent of the time. On the basis of these experiments, we propose that corneal allograft rejection requires a two-signal process for host sensitization.The K and O region (so-called classI) antigens present on all nucleated cells serve as one-signal. It appears from these experiments that la (classII) antigen-bearing cells (Langerhans cells and any passenger leukocytes) are a potent second signal. Further, when these cells are destroyed other minor histocompatibility antigens emerge as less potent second signals. Once the host generates cyctotoxic lymphocytes they appear to recognize and destroy any cells bearing non-host alloantigens.