The use of marginal liver donors can affect the outcomes of liver transplantation in patients with hepatitis C virus (HCV) infection. There are no firm conclusions about which donor criteria are important for allocation of high-risk grafts to recipients with HCV cirrhosis. We performed 120 consecutive liver transplantations for HCV infection between 1995 and 2005. Marginal donor criteria were considered to be: age >70 years, macrovesicular steatosis >30%, moderate-to-severe liver preservation injury, high inotropic drug dose (dopamine >15 microg/kg/min; epinephrine, norepinephrine, or dobutamine at any doses), peak serum sodium >155 mEq/L, any hypotensive episode <60 mm Hg and >1 hour, cold ischemia time >12 hours, ICU hospitalization >4 days, bilirubin >2 mg/dL, AST and/or ALT >200 UI/dL. Graft survival with donors showing these marginal criteria was compared with optimal donors using Kaplan-Meier analysis and the log-rank test. Independent predictors of survival were computed with the Cox proportional hazards model. Fifty-six grafts (46%) were lost during follow-up irrespective of the Model for End-Stage Liver Disease (MELD) scores of the recipients in each category. Upon univariate analysis, grafts with moderate-to-severe steatosis (P = .012), those with severe liver preservation injury (P = .007) and prolonged cold ischemia time (P = .0001) showed a dismal prognosis at 1, 3, and 5 years. Upon multivariate analysis, fat content (P = .0076; OR = 4.2) and cold ischemia time >12 hours (P = .034; OR = 7.001) were independent predictors of graft survival. Among HCV recipients, marginal liver donors worked similar to those from "good" donors, except for those with fatty livers >30%, especially when combined with a prolonged cold ischemia time.