Rationale Atractylodes macrocephala Koidz (Baizhu) is a valuable traditional Chinese medicine, and medicines of that type originating from Zhejiang province are the most famous and much more expensive than those from other regions. Driven by the great difference in prices, fraudulent labeling often occurs. In order to protect the interests of consumers, producers and honest traders, reliable techniques for the geographic authentication of Baizhu are needed urgently. Methods The stable isotope ratios of five light elements (C, N, H, O and S) in Baizhu samples originating from four provinces of China were determined with an elemental analyzer coupled to isotope ratio mass spectrometry, and the concentrations of 45 elements in these samples were measured using inductively coupled plasma mass spectrometry. Chemometric approaches including principal component analysis (PCA) and orthodox partial least squares discriminant analysis (OPLS-DA) were applied to the obtained data. Results The PCA results showed that the techniques enabled clear classification of the Baizhu samples into three clusters: A (Zhejiang province), B (Shaanxi province) and C (Hebei and Ahui provinces). Furthermore, OPLS-DA using 27 key variables provided 100% correct discrimination between samples originating from Zhejiang province and those from the other three provinces. Conclusions Stable isotope ratio and multielement analyses in combination with chemometric approaches showed great potential for the geographic authentication of Baizhu, providing a promising method for the control of fraudulent labeling that frequently occurs with traditional herbal medicines in China.
