Drug resistance reflects the evolution of tumors and represents the leading factor behind recurrence and death. Lenvatinib is the first-line therapy for hepatocellular carcinoma (HCC), but its effectiveness is limited by rapid development of resistance. Therefore, we aimed to identify lenvatinib resistance-related genes and assess their influence on prognosis and treatment response in HCC.The GSE186191 dataset served as the discovery cohort to identify lenvatinib resistance-related genes. A Venn diagram analysis delineated the intersection between lenvatinib resistance-related genes and prognostic-associated genes derived from The Cancer Genome Atlas (TCGA) database. The LASSO Cox regression model was implemented to construct a multigene signature in the TCGA cohort. A nomogram was built by integrating the TNM stage and our prognostic model. The gene signature and nomogram were further validated using HCC patients from the International Cancer Genome Consortium (ICGC) cohort. Mutation signatures, therapeutic response, functional enrichment, and immune profile analyses were performed in the two groups. Two lenvatinib-resistant (LR) HCC cells were established using a concentration gradient increment method. PFKFB4 expression was detected via qRT-PCR and western blot assay. The CCK-8 assay and flow cytometry were utilized to evaluate the proliferation and apoptosis of LR cells under different interventions.We developed a lenvatinib resistance-related gene signature (ALPK3, SLC2A2, CTSV, and PFKFB4), and demonstrated that's a precise, independent, and specific prognostic model for HCC patients. High-risk patients were characterized by a predisposition to TP53 mutations, aggressive tumor features, and treatment resistance. The risk score was significantly associated with immune cell infiltration, immune checkpoint expression, angiogenesis, and tumor stemness. PFKFB4 was overexpressed in LR cells, and its knockdown significantly enhances the antiproliferative and pro-apoptotic effects of lenvatinib on resistant cells.The lenvatinib resistance-related prognostic signature exhibits strong predictive power for prognosis in HCC patients and may serve as an effective tool for guiding treatment decisions. PFKFB4 promotes tumor progression and lenvatinib resistance, highlighting its potential as a novel therapeutic target for HCC.Not applicable.© 2025. The Author(s).