Graft function at 6 or 12 mo is positively correlated with renal transplant survival. The 36-mo results of a study that tested whether withdrawing cyclosporine (CsA) from a sirolimus (SRL)-CsA-steroid (ST) regimen would affect renal graft survival are reported. Eligible patients (n = 430) who were receiving SRL-CsA-ST were randomly assigned at 3 mo to remain on SRL-CsA-ST or to have CsA withdrawn (SRL-ST group). At 36 mo, the calculated GFR was significantly better with SRL-ST (47.3 versus 59.4 ml/min; P < 0.001) as was the slope of the GFR (-3.6 versus 0.8 ml/min; P < 0.001). This was accompanied by growing trend for improved graft survival in the SRL-ST group (85.1% versus 91.2%, P = 0.052 at 36 mo; 81.4% versus 91.2%, P = 0.015 in a cumulative data analysis up to 54 mo), despite numerically more biopsy-proven acute rejections after randomization (5.6% versus 10.2%; P = 0.107). Lipid parameters were similar between groups, whereas both systolic and diastolic BP were significantly lower in the SRL-ST group. Investigator-reported hypertension, abnormal kidney function, edema, hyperuricemia, hyperkalemia, gingival hyperplasia, and Herpes zoster occurred significantly more often in SRL-CsA-ST patients. Abnormal liver function test results, hypokalemia, thrombocytopenia, and abnormal healing were reported significantly more often with SRL-ST. The discontinuation rate was significantly higher for SRL-CsA-ST (48% versus 38%; P = 0.041). In conclusion, withdrawing CsA from a SRL-CsA-ST regimen at 3 mo after transplantation resulted in long-term benefits for renal transplant recipients.