Objective: This study aimed to develop a lipid metabolism-associated gene signature to stratify sepsis patients for prognostic risk and evaluate their immune function.Methods: The sepsis dataset GSE65682 from GEO was used. Lipid metabolism-associated genes were identified via GeneCards and intersection analysis. Key genes were selected by integrating Univariate Cox, LASSO, and Multivariate Cox regression analyses. Patients were stratified into high- and low-risk groups based on the median risk score. Prognostic performance was validated using Kaplan-Meier analysis and Receiver Operating Characteristic (ROC) curves. Immune heterogeneity was assessed via single-sample Gene Set Enrichment Analysis (ssGSEA), the CIBERSORT algorithm, and correlation network analysis. Results: A 9-gene prognostic signature (AHRR, CLN8, FASN, LSS, MED29, PAFAH1B1, PIP5K1C, TRIB3, UGCG) demonstrated robust predictive performance. High-risk patients exhibited poorer survival (KM test, P=6.75 × e-8; ROC AUC: 0.951) and enrichment in chemokine receptor signaling and parainflammation. In contrast, low-risk patients showed higher infiltration levels of tumor-infiltrating lymphocytes, type II interferon response, regulatory T cells, and macrophages. Immune network analysis revealed coordinated interactions: activated NK cells synergized with M1 macrophages (r=0.44) but antagonized resting NK cells (r=-0.62). Immune checkpoints CD86 and TNFSF4 were upregulated in low-risk patients, while CD200R1 was suppressed.Conclusion: The established lipid metabolism-based 9-gene signature effectively predicts sepsis outcomes, revealing significant immune differences between risk groups. It provides a potential tool for risk stratification and personalized clinical intervention.