This report extends the randomized, double-blind, multicenter trial at 14 centers in the United States that compared triple-therapy regimens containing mycophenolate mofetil (MMF), 2 or 3 g, with a similar regimen containing azathioprine (AZA) in recipients of cadaveric renal allografts. We investigated the continued long-term use of MMF with respect to graft and patient survival, graft function, and safety. All patients who received the study drug (MMF, 2 g, n = 165; MMF, 3 g, n = 166; AZA, n = 164) were included in the 3-year intent-to-treat evaluation of graft and patient survival. At 3 years posttransplantation, the weighted pairwise difference was 6.5%, and 95% confidence interval in graft and patient survival (ie, patients alive with a functioning graft) was -2.1 to 15.1 (P = 0.17) numerically in favor of MMF, 2 g, compared with AZA. Similar to the 1-year data, the principal adverse events were limited to the gastrointestinal and hematologic systems. Although cytomegalovirus (CMV) tissue-invasive disease occurred more often in the MMF-treated groups, most instances occurred during the first year posttransplantation. One patient in the AZA group and one patient in the MMF 2-g group developed lymphoma, whereas three patients in the MMF 3-g group developed lymphoma at 3 years posttransplantation. In conclusion, although the design of the study did not afford adequate statistical power to address survival end points, at 3 years posttransplantation, graft survival, patient survival, renal function, and safety were similar among the AZA, MMF 2-g, and MMF 3-g groups. There was an increased incidence of infection caused by invasive CMV and Aspergillus spp and mucormycosis seen in the MMF groups, but the long-term data confirm MMF is a safe and valuable addition to the range of drugs available to prevent rejection.