The long-term benefit of steroid withdrawal on patient and graft survival is unproven. Steroids were stopped within the first year after kidney transplantation in 223 consecutive low-risk patients initially treated with thymoglobulin and triple-drug therapy. The 15-year actuarial graft survival rate was 83.9%. Risk factors for graft loss were: proteinuria (hazard ratio [HR]: 6.96, P = 0.0003), creatinine >130 micromol/L (HR: 3.37, P = 0.01), recipient age <35 years (HR: 5.31, P = 0.001), and no mycophenolate mofetil (MMF) treatment (HR: 8.83, P = 0.04). Interestingly, recipient age and no MMF treatment were not risk factors in higher-risk patients in whom steroids were continued. The 2-year incidence of acute rejection following steroid withdrawal was 12.1%; the graft survival rate was lower in this group (71.1%). Our findings indicate that late steroid withdrawal results in excellent long-term outcome in most low-risk patients, but it should be attempted with caution in younger patients and when MMF is not used.