Abstract: OBJECTIVE To establish a method for simultaneous determination of amino acids, nucleosides and alkaloid1-deoxynojirimycin(DNJ) in mulberry leaves from different origins by UPLC-TQD, multivariate statistical analysis is employed to analysis discrepancies in compositions from these samples, to provide scientific basis for quality evaluation of mulberry leaf resources. METHODS Waters Acquity UPLC BEH Amide(100 mm×2.1 mm, 1.7 μm) column was employed, 0.2% formic acid-acetonitrile(A)-0.2% formic acid-water(B) was served as the mobile phase, gradient elution, flow rate was 0.4 mL·min^-1, column temperature was set at 35℃. Multiple reaction monitoring and positive ions mode was used to detect ions. Comprehensive evaluation of mulberry leaves from different areas were subjected to multivariate statistical analysis such as principal component analysis, etc. RESULTS The concentration of 22 components showed a good linear relationship consistent with the peak areas(r² ≥ 0.990 3). The RSD values of the precision, reproducibility and stability of the method were ≤ 7.5%. Moreover, the average sample recovery rate was 86.7%-106.4% and the corresponding RSD was 2.2%-8.9%. Mulberry leaves from different origins revealed great differences in amino acids, nucleosides and DNJ, combined with multivariate statistical analysis, these samples could be divided into two groups based on the discovered 8 most inhomogeneous componeous. CONCLUSION The method is accurate and reliable, combined with DNJ and 8 different components, mulberry leaves from Hubei and Jiangxi are deemed to be better.