Purpose: To compare a contact and a non-contact specular microscope in the determination of endothelial cell density. Subjects and methods: One hundred and twenty-one eyes from 70 patients who had undergone various degrees of photorefractive keratectomy for myopia were included. The endothelium was imaged by contact (Konan Clinical Specular Microscope) and non-contact (Topcon SP-1000) specular microscopy and the endothelial cell density estimated. Results: The average endothelial cell density achieved by the contact specular microscope was 3011+/-298 cells/mm(2) (mean+/-SD, n=121) and by the non-contact specular microscope 3015+/-265 cells/mm(2) (n=121), The difference in endothelial cell density between the contact and the non-contact specular microscope (contact minus non-contact) was -4+/-175 cells/mm(2) (t=0.26, 2p>0.05 in a paired t-test), The sampling error on the estimated endothelial cell density was 76 cells/mm(2) for the contact specular microscope and 74 cells/mm(2) for the noncontact specular microscope, \Conclusion: The average endothelial cell density and the precision of the measuring technique mere similar for the contact and the non-contact specular microscope. Furthermore, the endothelial cell densities estimated by the two instruments at various values of anterior central corneal refractive power and central corneal thickness were similar, The two instruments can be used interchangeably.