An ideal model to test methods of corneal storage for transplantation would simulate the environment of the grafted human cornea and predict the success of clinical corneal transplants (human to human). In this study, we tested such a model, the corneal xenograft (human to cat). Nine pairs of human corneas were transplanted into both eyes of nine recipient cats. One cornea of each pair was cryopreserved at -196 degrees C in 2.5 M dimethyl sulfoxide while the other was stored in preservative medium at 4 degrees C (control) for 6 +/- 2 (mean +/- SD) days before transplantation. One week after transplantation, the cats were euthanized and the eyes were examined. Three of the grafts (all cryopreserved) were clinical failures and showed no survival of donor corneal endothelial cells on scanning electron microscopy. The remaining six pairs of grafts were examined with a specular microscope and showed endothelial cell losses of 48 +/- 16% in cryopreserved and 8 +/- 16% in control corneas (p < 0.05). This survival is similar to survival in an earlier corneal perfusion model. The nine cryopreserved grafts were thicker than the control grafts, had fewer surviving keratocytes in the central stroma, and had more apoptotic central keratocytes (TUNEL assay). This failure rate in cryopreserved corneas clearly shows that this technique of cryopreservation was not adequate for clinical use. The corneal xenograft model can be used to study cellular survival and apoptosis in vivo after preservation as well as to test new methods of corneal preservation before initiating clinical trials.