HERAKLES was a 1-year randomized, multicenter trial. Patients were randomized at 3 months after kidney transplantation to remain on cyclosporine-based therapy, switch to everolimus without a calcineurin inhibitor (CNI), or switch to everolimus with low-exposure cyclosporine. Overall, 417 of 497 (83.9%) patients from the core study entered a 4-year extension study. The randomized regimen was continued to year 5 in 75.9%, 41.9% and 24.6% of patients in the standard-CNI, CNI-free and low-CNI groups, respectively. Adjusted estimated GFR at year 5 was significantly higher in the CNI-free group versus standard CNI (difference 7.2 mL/min/1.73 m , P < .001) or low CNI (difference 7.6 mL/min/1.73 m , P < .001). For patients who continued randomized therapy for 5 years, differences were 14.4 mL/min/1.73 m and 10.1 mL/min/1.73 m , respectively. Biopsy-proven acute rejection occurred during the 4-year extension study in 7.6%, 8.6%, and 9.0% of patients in the standard-CNI, CNI-free and low-CNI groups, respectively (P = .927). In conclusion, conversion to a CNI-free everolimus regimen 3 months after kidney transplantation improved long-term graft function, particularly in patients who continued the CNI-free regimen. Low CNI with everolimus did not improve renal function. Efficacy was comparable between groups but frequent immunosuppression changes should be taken into account.