Liver cancer (LC) is the fourth leading cause of cancer-related deaths worldwide. There is an urgent need to identify novel and reliable prognostic biomarkers for LC in order to improve patient outcomes. N6-methyladenosine (m6A) is the most common internal modification in eukaryotic mRNA and has been associated with various cancers, although its roles in the prognosis of LC remains to be elucidated. We analyzed the expression profiles of 15 m6A-related genes in the International Cancer Genome Consortium (ICGC) LIRI-JP dataset, and applied consensus clustering to stratify LC patients into two subgroups (Cluster 1 and Cluster 2). Cluster1 was significantly correlated to lower tumor stage and longer overall survival (OS). Gene set enrichment analysis showed that tumorigenic markers, including DNA repair, E2F targets, G2M checkpoint, and MYC targets V1, were enriched in Cluster2. We then constructed a prognostic risk model using three m6A-related genes that were identified as independent factors affecting OS. The nomogram based on the risk model score indicated good performance in predicting the 1-, 2-and 3-year survival of the LC patients. In conclusion, m6A-related genes are potential prognostic markers and therapeutic targets for LC.