Objective To study endothelial injury from a newly designed asymmetric double port Descemet Membrane Endothelial Keratoplasty (DMEK) injector, both ex-vivo and in clinical practice. Design Laboratory investigation with an interventional case series study. Method Sixteen rabbit endothelial rolls were tested for injection using a no-touch technique. For each pair of rolls, one endothelial graft underwent injection with a single port Pasteur pipette twice, wheras the other was injected with a novel asymmetric double port injector with a larger diameter entry port than the exit port also twice. Each graft was stained with 4-6-diamidino-2-phenylinidole dihydrochloride and was counted under a fluorescence-inverted microscope before and after injection. The proportion of graft injury was calculated and the differences were analyzed. Subsequently, six patients requiring DMEK underwent surgery using this novel insertion device and endothelial cell loss was calculated 3 months after the surgery. Results After injection, the mean proportion of endothelial cell survival with the single port pipette was 78.8% (n ¼ 8; SD: ± 20.9%), whereas the double port injector yielded a survival rate of 96.8% (n ¼ 8; SD: ±8.4%). This difference was statistically significant (P ¼ 0.008), representing less endothelial injury with the double port device. Early endothelial cell loss after 3 months in the DMEK patients was 26.1% (SD: ± 6.1%). Conclusion In our injection model, using a double port injector created significantly less endothelial cell damage than with the single port pipette. Clinically, this device yielded early endothelial cell loss comparable to that of the series performed by experienced DMEK surgeons.