To determine the benefits of long-term cyclosporine (CsA) immunosuppression, renal allograft recipients were randomly assigned to a protocol of either: CsA+azathioprine (Aza)+prednisone (TD), or to a protocol in which CsA was discontinued from the regimen of Aza+prednisone (CsA D/C). With a mean follow-up of nearly 7 years since transplantation, 30/47 (64%) CsA D/C and 27/45 (60%) TD had functioning allografts. Although long-term survivals were similar, hazards of the CsA D/C protocol were evident (40% rate of acute rejection following CsA D/C). Conversely, continued CsA in the TD protocol provided the opportunity for prednisone reduction, or even complete prednisone withdrawal in selected patients. A TD protocol which can provide equivalent long-term success, and eventually lower or omit prednisone, is preferable to a protocol of CsA D/C.