OBJECTIVE Renal insufficiency (RI) often coexists with hepatocellular carcinoma (HCC) and predicts a poor outcome in patients receiving transarterial chemoembolization (TACE). The modification of diet in renal disease (MDRD) and chronic kidney disease epidemiology collaboration (CKD-EPI) equations are used to provide estimated glomerular filtration rate (eGFR). This study aimed to determine a prognsotic renal surrogate for outcome prediction in HCC patients receiving TACE. METHODS A total of 684 patients with HCC undergoing TACE were retrospectively analyzed. The prognostic ability between the MDRD and CKD-EPI equation was compared by the Akaike information criterion (AIC). RESULTS Patients were categorized by eGFR into > 60, 30-60 and < 30 mL/min per 1.73 m(2) (equivalent to CKD stages 1-2, 3, and 4-5, respectively) groups. The eGFR generated by the MDRD equation had a better predictive accuracy by having a lower AIC score (3234.5) compared to the CKD-EPI equation (3236.7). The adjusted risk ratio for patients with eGFR 30-60 mL/min per 1.73 m(2) by the MDRD was 1.313 (P = 0.041) compared with patients with eGFR > 60 mL/min per 1.73 m(2) in the multivariate Cox model. The eGFR defined by the MDRD equation consistently had better prognostic ability when patients were stratified by the Child-Turcotte-Pugh score of 5 and > 5 and Cancer of the Liver Italian Program score 0 to 1 and > 1. CONCLUSIONS The eGFR according the MDRD equation may provide better prognostic accuracy than the CKD-EPI equation independent of liver functional reserve and tumor staging, and is a more feasible renal surrogate for outcome prediction in CKD stage 1-3 HCC patients receiving TACE.