Abstract: Based on the analysis of characteristics of mineral elements contents of foxtail millet from different producing areas, the geographical origin discrimination models were established in combination with chemometrics. In this study, major cultivars of foxtail millet from Longzhong area, Longdong area and Hexi area of Gansu Province were regarded as main research objects. The contents of 18 mineral elements in foxtail millet were determined by inductively coupled plasma mass spectrometry (ICP-MS), the data was analyzed by one way ANOVA, principal component analysis (PCA), orthogonal partial least squares-discriminant analysis (OPLS-DA), linear discriminant analysis (LDA) and hierarchical cluster analysis (HCA). The results showed that there was a significant difference for 13 mineral elements out 18 in foxtail millet from three producing areas (P<0.05), and mineral elements contents of foxtail millet showed distinct regional distribution characteristic. Correlation analysis showed there were significant relationships between 18 mineral elements. Four principal components were extracted after PCA, and the cumulative contribution ratio of the four components was 75.82%. Correct discrimination rates of LDA and OPLS-DA models were both 100%, which could be used for the geographical origin discriminantion of foxtail millet in different producing areas of Gansu Province, and six characteristic elements (V、Fe、Cu、Cd、Se、Pb) of foxtail millet from different producing areas were screened by OPLS-DA models. HCA regarding the characteristic elements as variables could classify the foxtail millet into different categories, which was consistent with their geographical origins. The research suggests that the discrimination model based on mineral element contents can effectively identify foxtail millet from different producing areas, which provides a scientific basis for origin traceability discrimination and quality control of foxtail millet.