Abstract: To promote metabolomics technology in tobacco research, to standardize tobacco metabolomics data acquisition process, and to improve data's comparability and usability, a tobacco metabolomics workflow was established to cover fresh leave collection and mass spectrometry analysis. Fresh tobacco leaves were quickly wrapped using tin foil after picking, and then frozen in liquid nitrogen. Fully frozen tobacco leaves were embedded with dry ice and transferred to laboratories, and then grounded into powder after being lyophilized in vacuum for metabolite extraction. Tobacco metabolomics data were collected with nine different methods based mainly on GC-MS and LC-MS, including six targeted methods (on phytochromes, free amino acids, organic acids, alkaloids, polyphenols and terpenoids) and three untargeted methods (lipids, all components analysis based on GC-MS or LC-MS). Fresh tenth tobacco leaves (counted from bottom to top) sampled from different cultivars and planting areas at different growth stages were analyzed using the established methods. The results showed that:1) Up to five hundred metabolites including mostly primary and secondary metabolites were qualified and quantified, of which 101 metabolites were absolutely quantified. 2) The metabolome of fresh tobacco leaves, indicated by principal component analysis (PCA) method, differed greatly at different growth stages, e.g. the contents of most amines (including free amino acids) in tobacco leaves gradually decreased during growing. At the same growth stage, ecological differences were greater than variety differences in the metabolome of fresh tobacco leaves. This method is stable, reliable and suitable for tobacco metabolomics research with the characteristics of high repeatability, rich data and broad coverage of metabolites in tobacco leaves.