The transcorneal penetration of cyclosporine A has been determined from each of three vehicles across isolated cornea into simulated aqueous humor containing either 50 mg % protein (0.5 mg/ml; as found in a normal eye) or 5000 mg % protein (50 mg/ml; as found in an inflamed eye). Cyclosporine entered the corneal epithelium and stroma/ endothelium as well as passed through the cornea from an alpha cyclodextrin vehicle. Entry into the epithelium and stroma/endothelium occurred from an ointment vehicle with limited detectable anterior chamber penetration using 50 mg % protein solution in the anterior chamber. From corn oil vehicle, cyclosporine penetrated across the cornea with a permeability equal to that of alpha cyclodextrin vehicle. The concentration of cyclosporine in both corn oil and ointment vehicles is 8 times greater than that in alpha cyclodextrin vehicle resulting in a flux from corn oil vehicle about 7 or 8 times greater than that seen after alpha cyclodextrin vehicle. The amounts retained in the cornea, however, were relatively low after corn oil compared to cyclodextrin. The penetration of cyclosporine from either the cyclodextrin vehicle or ointment was at least doubled in the presence of 5000 mg % protein in the simulated aqueous humor relative to that seen in 50 mg % protein. This data indicates that the (presumed) absorption and binding of drug by the excess protein in the simulated aqueous humor may have removed free cyclosporine from the solution and sustained a high concentration gradient of free solute across the cornea. This occurs despite the proven binding of cyclosporine to the tubing leading from the corneal chamber to the collection vial, although the influence of albumin upon this binding is relatively small. The vehicle governs the release an penetration of cyclosporine to and through the cornea, with corn oil and cyclodextrin giving the same corneal permeability to cyclosporine but since corn oil can contain a greater drug concentration it provides a greater flux of the drug.