The purpose of this study was to determine the pharmacokinetics of cyclosporine in the cornea and aqueous humor following topical application in collagen particles and to determine the capacity of cyclosporine in collagen particles to suppress corneal allograft rejection. Cyclosporine was prepared for topical application in three vehicles: collagen particles suspended in methylcellulose, collagen shields, and corn oil. Rabbit eyes were treated with a single application of one formulation. Corneas, aqueous humor samples, and blood samples were collected at 2, 4, 8, 12, and 24 hours. Peak concentrations of drug were reached between 2 and 4 hours in the cornea and between 4 and 8 hours in the aqueous humor. No drug was found in the blood. Collagen particles and shields produced significantly higher concentrations of cyclosporine in the aqueous humor and cornea, compared with corn oil drops, at virtually all time points. In the second part of the study, rabbits with vascularized corneas underwent corneal transplantation and were treated with one of the formulations beginning either at the time of transplantation (prophylactically) or at the first sign of an allograft immune reaction (delayed treatment). Cyclosporine delivered in collagen particles was effective in preventing primary corneal allograft rejection in nearly all of the grafts (96%) treated prophylactically, and in 67% of the grafts treated when signs of the first stage (grade 1) of immune graft reaction were observed. Grafts treated with cyclosporine in collagen shields showed similar survival rates (87% and 64%, respectively), whereas only 40% of the grafts treated prophylactically and only one (18%) receiving delayed treatment with the corn oil preparation survived. These results demonstrate the efficacy of collagen particles for delivery of hydrophobic drugs to the ocular surface and the potential for preventing corneal graft rejection with topically applied cyclosporine in a collagen-particle vehicle.