High-dose systemic cyclosporine (CyA) administration frequently results in severe side effects. To evaluate a means of limiting the adverse effects of CyA, we implanted CyA-collagen matrices (0.2 or 1 mg/kg/day released) around the cardiac homografts at the time of rat heterotopic (neck) heart transplantation. Control animals received empty collagen (nondrug) matrix implants. A fourth group received CyA matrix (1 mg/kg/day released), implanted in a distal subdermal leg pouch at the time of heart transplantation. Rejection was determined by the absence of contraction in the transplanted heart. No animal received any other immunosuppression. Parallel groups of animals had whole blood, heart, and kidney CyA levels measured on the sixth posttransplant day. Local immunosuppression with high-dose CyA in a controlled-release matrix resulted in a significant survival advantage (mean survival time, 17.1 days; control, 6.9 days; p less than 0.001). The lower dose of CyA also demonstrated significant survival benefits (10.1 days), with clinically negligible blood CyA levels and very low kidney CyA levels. Both doses of epicardial local release CyA were well absorbed locally, resulting in very high CyA levels in cardiac tissue. Local immunotherapy of transplanted hearts with CyA was shown to be an effective means of preventing rejection. If this technology can be developed, this approach may prove advantageous clinically, both in extending transplantation and in minimizing systemic side effects of immunosuppression.