Purpose: To evaluate the effect of the location of the corneoscleral tunnel incision as well as preoperative and intraoperative parameters on total and localized endothelial cell loss. Setting: Department of Ophthalmology, Humboldt-University of Berlin, Berlin, Germany. Methods: Fifty consecutive patients scheduled for routine cataract surgery were selected prospectively for this clinical trial. Preoperatively, the axial length, anterior chamber depth, lens thickness, and astigmatism were measured. Phacoemulsification time and relative energy as well as total surgical time were recorded. With a specular microscope, endothelial cell counts were determined centrally, superiorly, and temporally preoperatively and 6 weeks and 6 and 12 months postoperatively. Results: After 12 months, the mean overall central endothelial cell loss in all eyes was 8.5%. The mean endothelial cell loss was 11.9% in the lateral quadrant and 11.4% in the superior quadrant. There were no significant differences between superior and temporal surgical approaches in intraoperative parameters of phacoemulsification time, relative intensity of phacoemulsification, and surgical time. There were no significant differences in central endothelial cell loss or in the area localized in the quadrant of the positions of the corneal surgical site. The only risk factors found significant for higher endothelial cell loss were shorter axial length and longer phacoemulsification time. Conclusions: The location of corneoscleral incisions for phacoemulsification can be chosen according to the preoperative astigmatism without inducing additional adverse effects on the comeal endothelium. Shorter eyes have a significantly higher risk for greater endothelial cell loss. (C) 2000 ASCRS and ESCRS.