To explore the metabolic differences between green and jasmine tea, untargeted metabolomics analysis of green tea and jasmine tea samples from three different regions was performed using ultra-high-performance liquid chromatography–high resolution mass spectrometry (UHPLC-HRMS). A total of 177 metabolites were detected and Kruskal-Wallis tests were performed, leading to the selection of 59 metabolites that were then classified into three groups through Pearson cluster analysis. Metabolic pathway enrichment analysis results indicated that metabolic differences mainly focused on polyphenol, flavone, and amino acid metabolism. The results of partial least squares-discriminant analysis indicated that jasmine tea and green tea can be clearly distinguished by score scatter plots. After filtering based on Variable Importance in Projection values, 28 differential metabolites were retained. Correlation analysis between the total polyphenol and flavonoid contents and differential metabolites revealed that γ-aminobutyric acid and lyso-phosphatidylcholine 18:0 exhibited the strongest positive and negative correlation, respectively. Analysis of green tea and jasmine tea from different regions identified 10 and 55 differential metabolites, respectively. Among them, Anhui, Jiangxi, and Shaanxi green tea had the highest 12-oxo-phytodienoic acid, jasmone, and phloroglucinol contents. Moreover, Anhui, Guangxi, and Sichuan jasmine tea had the highest acetophenone, lyso-diacylglyceryltrimethylhomoserine 16:0, and 2-hexenal contents. This study provides reference for research on the formation of tea flavors and identification of tea from different producing areas.