Metabolic pathways exert a significant influence on the onset and progression of cancer. Public data on hepatocellular carcinoma (HCC) patients were obtained from The Cancer Genome Atlas (TCGA) and International Cancer Genome Consortium (ICGC) databases. Analysis was performed in R software using different R packages. Here, we integrated the data from multiple independent HCC cohorts, including TCGA‐LIHC, ICGC‐FR and ICGC‐JP. Then, the enrichment score of 21 metabolism‐related pathways was quantified using the ssGSEA algorithm. Next, univariate Cox regression analysis was applied to identify the metabolic terms with significant correlation to patient survival. Finally, a prognosis model based on linoleic acid metabolism, sphingolipid metabolism and regulation of lipolysis in adipocytes was established, which showed good performance in predicting patients' survival. Furthermore, we conducted a biological enrichment analysis to delineate the biological disparities between high‐ and low‐risk patients. Notably, we discerned differences in the microenvironments between these two patient groups. We also found that low‐risk patients could potentially respond better to immunotherapy. Drug sensitivity analysis suggested that low‐risk patients are more susceptible to bexarotene and erlotinib, yet exhibit resistance to ATRA and bleomycin. Furthermore, through the use of LASSO logistic regression analysis, we identified 19 characteristic genes, which could robustly indicate the risk groups. Our research underscores the role of linoleic acid metabolism, sphingolipid metabolism and the regulation of lipolysis in adipocytes in HCC, pointing towards potential avenues for future research.