The immunotherapy of endometrial cancer (EC) has gradually attracted attention and metabolic reprogramming is associate with tumor immune infiltration. Our goal was to use proteome analysis to examine the role of immune-related metabolic genes (IRMGs) in EC. Data-independent acquisition mass spectrometry (DIA-MS) was performed on 20 EC patients, consisting of 10 high-grade and 10 low-grade cancer tissues. IRMGs were screened using Spearman correlation, and an immune-related metabolic prognosis signature (IRMPS) was constructed based on the Cancer Genome Atlas-Uterine Corpus Endometrioid Carcinoma (TCGA-UCEC) cohort using the least absolute shrinkage and selection operator (LASSO) regression analysis. We also investigated differences between different risk groups in terms of prognostic value, clinical potency, immune characteristics and therapy response. In total, 285 differentially expressed genes (DEGs) were acquired via DIA-MS. Subsequently, metabolic-DEGs and immune-DEGs were analyzed by Spearman correlation to identify 41 IRMGs. Finally, seven IRMGs, including NADH dehydrogenase (ubiquinone) 1 alpha subcomplex subunit 2 (NDUFA2), AMPK-alpha2 (PRKAA2), syntaxin binding protein 1 (STXBP1), NADH dehydrogenase (ubiquinone) 1 beta subcomplex subunit 9 (NDUFB9), ribosomal protein S27-like (RPS27L), lysolecithin acyltransferase 2 (LPCAT2) and uridine monophosphate synthetase (UMPS) were identified to establish a prognosis signature. The risk score was determined as an independent prognostic indicator, and patients in the IRMPS-high group was strong linked with adverse prognosis for EC. Additionally, IRMPS was closely related with tumor immune infiltration. Notably, the IRMPS-low group had better immune checkpoint inhibitors (ICI) treatment response and more sensitive to chemotherapy drugs. In conclusion, IRMPS can serve as a precise prognostic tool to guide the personalized treatment of EC patients.