We conducted a multicenter, randomized, controlled trial to investigate the effect of long-term oral supplementation with branched-chain amino acids (BCAA) on the event-free survival in 622 patients with decompensated cirrhosis. In the present study, the development of liver cancer was analyzed as an endpoint in particular. Subjects received either treatment with BCAA at 12g/day or dietary therapy containing the matched daily energy and protein intake. A Cox regression analysis was carried out to estimate the hazard ratios for different background factors stratified by treatment group. Liver cancer was noted in 89 patients. The risk for liver cancer was significantly higher for males, patients with concurrent diabetes mellitus, patients with an alpha-fetoprotein (AFP) level of 20ng/mL or higher, patients with higher body mass index (BMI), and patients with lower serum albumin levels. When the BCAA group and the diet group were compared for factors that interacted with the treatment arms, the risk for liver cancer was significantly reduced in the BCAA group with a BMI of 25 or higher and with an AFP level of 20ng/mL or higher. Oral supplemental treatment with BCAA may reduce the risk of liver cancer in cirrhotic patients with these specific factors.