Bipolar disorder (BD) is a debilitating mental disorder with complex clinical manifestations and low diagnostic accuracy. Depressive episodes are most common in the course of BD with high comorbidity and suicide rates, which present greater clinical challenges than mania and hypomania episodes. However, there are no objective biomarkers for bipolar depression. The aim of this study was to detect urinary metabolite biomarkers that could be useful for the diagnosis of bipolar depression. Nuclear magnetic resonance spectroscopy was used to profile urine samples of patients with bipolar depression (n = 37) and healthy volunteers (n = 48). Data were analyzed using Orthogonal Partial Least Square Discriminant Analysis and t-test. Differential metabolites were identified (VIP > 1 and p < 0.05), and further analyzed using Metabo Analyst 3.0 to identify associated metabolic pathways. In total, we identified seven metabolites differentially expressed in patients with BD and healthy controls. Compared with healthy group, the levels of betaine, glycerol, hippuric acid, indole sulfate, trimethylamine oxide, and urea in urine samples of BD patients were significantly higher, while the level of inositol was significantly lower. Most of these small molecules are related to lipid metabolism and gut microbiota metabolism. These differential metabolites could provide critical insight into the pathological mechanisms of bipolar depression. The results of this study provide a meaningful reference for similar and further studies in the future.Copyright © 2020. Published by Elsevier B.V.