This study aimed to compare the effects of reduced glutathione (GSH) with those of S-(1,2-dicarboxyethyl)glutathione (DCE-GS) (in rabbits and humans), and different concentrations of the latter (in humans), on corneal endothelial permeability when added to solutions bathing the isolated cornea. Inulin/dextran permeability was determined from stromal- to endothelial-facing surfaces of de-epithelialized corneas. The bathing solution was modified Opeguard®-MA (MOMA), an ocular irrigating solution, to which either GSH or another intrinsic tripeptide, DCE-GS, was added. Paired corneas were used to compare either different combinations of GSH with DCE-GS (rabbit or human) or various concentrations of DCE-GS from 0.25 to 2.0 mM (human). Endothelial cyclic AMP levels were determined in cultured rabbit cells. MOMA alone resulted in approximately the same permeability as MOMA + 0.3 mM GSH while the use of 2 mM DCE-GS significantly reduced rabbit (40% maximum, p < 0.00001) and human (30% maximum, p < 0.01) corneal permeability. Human corneal endothelial permeability remained reduced through a range of concentrations of DCE-GS from 0.25 to 2.0 mM DCE-GS. Tissue-cultured rabbit corneal endothelium showed an increase in cyclic AMP after DCE-GS or GSH. DCE-GS potentially offers a viable alternative to GSH for inclusion in ocular irrigating or corneal preservative solutions since it maintains human corneal endothelial permeability at a lower, stable value relative to non-DCE-GS-containing solutions.