Though conventional corneal transplantation has achieved great success, it still has several drawbacks including limited availability of donor corneas, recurrent allograft rejection, and subsequent graft failure in certain cases. Reconstructing clinically usable corneas by applying the technology of regenerative medicine can offer a solution to these problems, as well as making corneal transplantation a non-emergency surgery and enabling the usage of banked corneal cells. In the present study, we focused on corneal endothelium that is critical for corneal transparency and investigated the reconstruction of cornea utilizing cultured human corneal endothelial cells (HCECs). We succeeded in steadily culturing HCECs by using culture dishes pre-coated with extracellular matrix produced by calf corneal endothelial cells and culture media that contained basic fibroblast growth factor and fetal bovine serum. We performed the following analysis utilizing these cultured HCECs. The older the donor was, the more frequently large senescent cells appeared in the passaged HCECs. The telomeres of HCECs were measured as terminal restriction fragments (TRF) by Southern blotting. HCECs, in vivo from donors in their seventies had a long TRFs of over 12 kilobases. Passaging shortened the TRFs but there was no difference in TRFs among donors of various ages. These results indicated that shortening of telomere length is not related to senescence of HCECs. We investigated the role of advanced glycation end products (AGEs) in the senescence of in vivo HCECs. The results indicated that AGE-protein in the aqueous humor is endocytosed into HCECs via AGE receptors expressed on the surface of HCECs and damages HCECs by producing reactive oxygen species and inducing apoptosis, suggesting that AGEs, at least partly, cause the senescence of HECEs. HCECs were cultured using adult human serum instead of bovine serum to get rid of bovine material that can be infected with prions. Primary and passage culture of HCECs was possible using adult human serum. We reconstructed the cornea using cultured HCECs and human corneal stroma. The corneal stroma, on which the cell suspension of HCECs was poured, was mildly centrifuged to enhance the HCECs attachment to the stroma. The cell density of HCECs on the reconstructed cornea reached 2,500 cells/mm2. The pump function of the reconstructed cornea was measured with an Ussing chamber. The potential difference in the reconstructed cornea and normal cornea was 0.30 mV and 0.40 mV, respectively; indicating that the pump function of the reconstructed cornea is 75% of that of the normal cornea. The reconstructed cornea was transplanted to a rabbit eye and stayed transparent for 6 months after the operation. Fluorescein labeled cultured HCECs remained on the graft 1 month after the transplantation, indicating that transplanted HCECs contributed to the transparency of the graft. The possibility of using artificial stroma or porcine corneal stroma as a carrier of cultured HCECs was investigated. The artificial stroma made of alkaline-treated collagen could not be sutured but showed good transparency, biocompatibility, and cell-attachability. Porcine corneal stroma, expressing little xeno-sugar antigen alpha-gal epitope, induced no super acute rejection but mild cellular rejection when transplanted in the cornea of animals possessing natural antibody to alpha-gal epitope. The cornea reconstructed with porcine corneal stroma and HCECs had an average cell density of 1721/mm2 and had approximately 60% of the pump function of a normal cornea. As new technologies in corneal transplantation, the application of self immature cells and the direct delivery of cultured HCECs into the anterior chamber were investigated. Part of rat mononuclear cells that were obtained from the bone marrow and injected into the rat anterior chamber transformed into corneal endothelium-like cells, suggesting that self immature cells can transform into corneal endothelial cells. Cultured rabbit corneal endothelial cells that endocytosed iron were injected into the anterior chamber of rabbits whose corneal endothelium was cryo-injured, and were pulled to Descemet's membrane by putting a magnet on the eyelid. In these rabbits, corneal edema decreased more quickly than in the control group and no intraocular pressure rise was observed during 8 weeks after the operation, suggesting that the direct delivery of cultured HCECs into the anterior chamber can be an alternative method of choice. The following obstacles should be addressed to make the transplantation of cultured corneal endothelial cells clinically applicable. 1. To reconstruct a cornea that is the same as or superior to the normal cornea, more innovation is necessary in the method of culturing and seeding HCECs. We should consider utilizing HCECs obtained from fetuses after clearing ethical issues. Moreover, we need to develop a method to enhance the cell density and the cell functions. 2. Porcine corneal stroma is promising as a carrier of HCECs instead of human corneal stroma, which is in very limited supply. The usefulness of porcine corneal stroma acellularized to prevent retrovirus infection should be evaluated. 3. To make the self immature cells applicable to corneal transplantation, we should elucidate the corneal endothelial cell specific markers and the factors that are necessary to induce self immature cells to become corneal endothelial cells. 4. The direct delivery of cultured HCECs into the anterior chamber can be an alternative method of choice when its long-term safety is confirmed.