Autoimmune disease is a condition that arises from abnormal immune responses to self-antigen, which leads to damage of tissues or dysfunction of biological system. Graves’ disease (GD) is an autoimmune disorder that affects thyroid, and frequently results in hyperthyroidism. It often develops to an enlarged thyroid and eye bulging, Graves' ophthalmopathy (GO) with 25% frequency of GD. Since there is no specific molecular indicator for the disease, biomarker based on blood metabolite can be an effective and powerful tool for diagnosis, and suggest putative therapeutic target. In this study, a gas chromatography coupled time-of-flight mass spectrometry was employed to characterize primary metabolic profiles from plasma of GD (n=20), GO (n=27), and healthy control (n=32). Multivariate statistical analysis were applied to identify the metabolic phenotype of Graves’ disease compared to health control. The most significant alteration was found in glycolysis and branched chain amino acids. The subsequent biomarker model consisting of 1,5-anhydroglucitol, proline, glucose and pyruvate showed excellent discrimination power. Area under the curve (AUC) values by receiver operating characteristic (ROC) analysis were 0.94 (control vs others), 0.87(GO vs others) and 0.86(GD vs others), respectively. In addition, lipid profiling revealed the dramatic alteration of sphingolipids and phospholipids in the disease, which included oleoyl-LPA and sphingosine 1-phosphate.