Background: Patients with hepatocellular carcinoma (HCC) may develop end-stage renal disease and receive dialysis, but the impact of dialysis on the prognosis is unclear. This study aimed to evaluate the outcome of dialysis HCC patients and the prognostic role of albumin-bilirubin (ALBI) grade in these patients. Methods: Among the consecutive 3,794 HCC patients between 2002-2017, 43 patients undergoing dialysis, and 129 age, sex-matched controls were analyzed. Multivariate Cox hazards model was used to identify independent prognostic predictors. Results: Dialysis patients had decreased overall survival when compared with non-dialysis patients (n=3,751) and matched controls (n=129; each P=0.004). Patients with ALBI grade 1 had the best survival in the pooled cohort of dialysis and matched controls (n=172). In the Cox model, total tumor volume >33 cm 3 [hazard ratio (HR): 6.763, P<0.001], presence of ascites (HR: 6.168, P<0.001), dialysis duration less than 24 months (HR: 3.144, P=0.006), diabetes-related dialysis (HR: 9.366, P=0.001) and non-curative treatments (HR: 9.220, P<0.001) were poor prognosis factors associated with increase mortality among dialysis patients. Of the 9currently-used HCC staging systems, the CLIP score was the optimal cancer staging for dialysis patients. Conclusions: Patients receiving dialysis had decreased overall survival compared with non-dialysis patients. Longer duration of dialysis, non-diabetes related dialysis, absence of ascites, and curative treatments were associated with improved survival in these patients. The ALBI grade is a feasible prognostic model to evaluate liver functional reserve, and the CLIP model is the best staging system for dialysis patients with HCC.