Recently, it has been a big issue to distinguish the dried roots of Cynanchum wilfordii and Cynanchum auriculatum in Korean herbal medicine market. Although C. wilfordii and C. auriculatum have similar morphology, the types and quantities of metabolites may differ depending on the species. Thus, in this study, UPLC-QTOF/MS based metabolomics was applied to discriminate the roots of C. wilfordii and C. auriculatum. In the optimal LC/MS conditions, 64 known metabolites were analyzed in the two species. PCA and PLS-DA of metabolic profile data was able to differentiate between C. wilfordii and C. auriculatum. Furthermore, OPLS-DA and S-plot were applied to find the potential biomarkers for the discrimination of C. wilfordii and C. auriculatum. Finally, 4 known and 10 unknown metabolites were determined as the biomarkers, and their repeatability and reliability were also validated. This indicated the metabolite profiling is a robust approach to find discriminating biomarkers of C. wilfordii and C. auriculatum.