Helwingia japonica (HJ), a traditional medicinal plant, is commonly used for the treatment of dysentery, blood in the stool, and scald burns. Three major HJ species, Helwingia japonica (Thunb.) Dietr. (QJY), Helwingia himalaica Hook. f. et Thoms. ex C. B. Clarke, and Helwingia chinensis Batal., share great similarities in both morphology and chemical constituents. The discrimination of medicinal plants directly affects their pharmacological and clinical effects. Here, we solved the taxonomy uncertainty of these three HJ species and explored the discrimination and study of other traditional medicines (TMs). First, the anti-inflammatory effects of the three HJ species were compared using lipopolysaccharide (LPS)-induced inflammatory responses in mouse leukemia cells of monocyte macrophage (RAW) 264.7 cells. Then, plant metabolomics were performed in 48 batches of samples to discover chemical markers for discriminating different HJ species. Finally, network pharmacology was applied to explore the linkages among constituents, targets, and signaling pathways. In vitro experiments showed that the QJY exhibited the most potential anti-inflammatory activities. Meanwhile, 172 compounds were tentatively identified and eight metabolites with higher relative content in QJY were designated as chemical markers to distinguish QJY and the other two species. According to the property of absorbed in vivo, threonic acid, arginine, and tyrosine were selected to construct a component-target-pathway network. The network pharmacology analysis confirmed that the chemotaxonomy differentiation was consistent with the bioactive assessment. The present study demonstrates that bioactivity evaluation integrated with plant metabolomics and network pharmacology could be used as an effective approach to discriminate different TMs and discover the active compounds.