Post-transplant outcomes for acute liver failure (ALF) are unsatisfactory, and there are debates about the most suitable type of graft. Given the critical shortage of donor organs, accurate assessment of post-transplant outcome in ALF patients is crucial to avoid a futile liver transplantation (LT). A database of 160 consecutive adult ALF patients who underwent primary LT between 2000 and 2009 in a tertiary LT center was analyzed. The most common causes of ALF were hepatitis B virus infection (30%) and herbal/folk medicine use (30%). Thirty-six (22.5%) and 124 (77.5%) patients underwent deceased-donor LT (DDLT) and adult-to-adult living-donor LT (LDLT), respectively. During a median follow-up period of 38 (range 1-132) months, the DDLT and LDLT groups showed similar patient (P=0.99) and graft (P=0.97) survival rates. The overall 1- and 3-year patient survival rates were 78.8 and 74.6%, respectively. Five predictors of patient survival were identified by bootstrapping and multivariate analysis: vasopressor requirement, estimated glomerular filtration rate, serum sodium concentration, recipient age, and donor age, at the time of transplant. By summing scores weighted in each of these predictor categories, we designed a prognostic scoring system (scores from -2 to 20) that estimated 1-year post-transplant mortality from 0 to 100% (c statistic 0.79). Long-term outcomes after LDLT and DDLT were comparable in adult patients with ALF. A simple prognostic scoring system that includes 5 predictive variables at the time of LT may help estimate post-transplant survival in ALF patients, regardless of the type of transplant.