Retrograde arterial flush (RGAF) of liver grafts, which prevents arterial intima injuries in living donor liver transplantation (LDLT), seemed to reduce post-transplantational cholestasis by our preliminary study. The aim of this study was to further confirm this effect of RGAF of grafts in LDLT by a prospective randomized study. The enrolled cases were randomly divided into two groups, RGAF (portal flush plus RGAF) and non-RGAF (portal flush only). The outcome measures included intra-operative hemodynamic changes, one-month post-transplantational liver function tests and acute cellular rejection, vascular and biliary complications, the length of postoperative hospital stay, and graft and patient survivals. The significant findings of results were lower postoperative serum bilirubin in the RGAF group until three wk after transplantation (days 1, 3, 7, 14, and 21 serum total bilirubin, mg/dL: 5.82±5.46, 3.95±4.28, 4.87±5.01, 3.62±5.32, and 3.63±6.91 vs. 8.41±5.49, 7.47±5.61, 10.0±10.1, 9.06±11.6, and 6.02±9.84, respectively, in RGAF and non-RGAF, p=0.017, 0.002, 0.014, 0.011, 0.109) and shorter postoperative hospital stay in the RGAF group (31.1±17.6 vs. 44.8±32.3 d, respectively, in RGAF and non-RGAF, p=0.035). In conclusion, the RGAF of liver grafts on a back table may ameliorate postoperative functional cholestasis in LDLT, which possibly reduces postoperative hospital stay.