Description:Thymoma is a rare malignant thymic epithelial tumor (TET) with an incidence of about 2 to 4 per million. In this study, we performed nontargeted lipidomic profiling study of tissue samples from 39 patients with thymoma and thymic hyperplasia using UPLC-HRMS. OPLS-DA classification models enabled discrimination of tumor from adjacent noncancerous tissue (ANT) and distant noncancerous tissue (DNT) samples in thymoma with good accuracy, sensitivity, and specificity using 5 discriminant compounds. The discriminant feature panels were verified by external validation sets, revealing characteristic lipid signatures associated with thymoma and showing prospects for entering the next evaluation stage of biomarker discovery for thymoma detection, giving hints to potential treatment targets for the disease.