Endometrial cancer (EC) is a malignant tumor of the female reproductive tract. Due to its rapid growth and invasiveness, EC is currently the only gynecological neoplasm with rising incidence and mortality rates. It is of great significance to explore the metabolomics signature of stage I and II EC for the diagnosis and treatment. A mass spectrometry-based untargeted metabolomics approach was used to explore preoperative serum metabolites in the normal and stage I and II EC patients. The metabolites were mapped to the Ingenuity pathway analysis (IPA) database to determine the potential biomarkers and metabolic pathways that differ between EC patients and healthy controls. The top analysis-ready molecules of upregulated D-glucose thiamine and downregulated cholesterol, arachidonic acid, palmitic acid, oleic acid, stearic acid, linoleic acid may be the most related metabolites. These potential biomarkers have essential functions in regulating vital metabolic pathways associated with stage I and II EC. Additionally, our pathway analysis revealed five significantly related pathways according to the metabolite differentials. Finally, the disease and function prediction of the initial pathway analysis suggested that small molecule biochemistry, lipid metabolism, and organismal injury and abnormalities were associated with EC cases. Over 25 metabolites were differentially expressed in stage I and II EC. In addition, the six most significant metabolites were related to stage I and stage II EC cases. Ingenuity pathway analysis revealed potential biomarkers and metabolic pathways revolved to EC. In this paper, candidate endogenous biomarkers were defined as the basis for disease diagnosis and individualized treatment monitoring and revealed the mechanism of EC occurrence and development.