Background: Long-term use of calcineurin inhibitors (CNI; cyclosporine and tacrolimus [TAC]) is associated with chronic renal dysfunction, a significant cause of morbidity in liver transplant recipients (LTxR). Early everolimus (EVR) facilitated reduction or elimination of CNI may reduce CNI-associated nephrotoxicity in de novo LTxR. Methods: H2304 (NCT00622869) is a 24-month (M), randomised, multicenter, open-label study in 719 de novo LTxR. After 30 days with TAC±mycophenolic acid, LTxR were randomised 1:1:1 into 3 arms comparing (1) EVR (C0 3-8ng/mL)+ reduced TAC (C0 3-5ng/mL; EVR+rTAC; N=245) or (2) EVR (C0 6-10ng/mL)+TAC withdrawal (TAC- WD; N=231) at M4 to (3) standard TAC (C0 6-10ng/mL; TAC-C; N=243); all with steroids. Main endpoints at M12 included the composite efficacy failure rate of treated BPAR, graft loss or death and evolution of renal function (RF) from randomisation (RND) to M12 measured as eGFR (MDRD4; mL/min/1.73m 2 ). Results: Enrollment in TAC-WD arm was stopped early due to high AR rate clustered around time of TAC elimination; statistical comparison is only presented for EVR+rTAC vs TAC-C. At M12, the composite efficacy failure rate in EVR+rTAC arm was comparable to TAC-C (6.7 vs 9.7%, p=0.230); however, was higher in TAC-WD (24.2%). Evolution of RF from RND to M12 was superior for EVR+rTAC over TAC-C (adjusted mean difference [SE] in eGFR: +8.50mL/min/1.73m 2 [2.12], 97.5% CI [3.74, 13.27];