Abstract: 【Objective】 This project aims to analyze the contents of 15 mineral elements in wolfberry and compare the differences in mineral element content of fruit from different years and geographical origins. The purpose is to establish a authenticity identification model based on multiple mineral element analysis and to explore the stability and reliability of authenticity identification methods for origin. 【Methods】 A total of 264 wolfberry fruit samples were collected from Jinghe County of Xinjiang and Zhongning County of Ningxia for 3 consecutive years (2020-2022). The contents of 15 mineral elements such as Na, Mo, Mn, Rb, Mg, Ca, Cu, Sr, Cr, Ni, K, Co, Ba, Fe and Zn were analyzed by inductively coupled plasma mass spectrometer, and the origin was identified by analysis of variance and orthogonal partial least square discriminant analysis (OPLS-DA). 【Results】 There were significant differences in 7 mineral elements of Na, Cu, Sr, Mg, Mo, Rb and Mn between Jinghe and Zhongning (P<0.05). 13 elements of Na, Mo, Mn, Rb, Mg, Ca, Cu, Sr, Cr, K, Co, Ba, Zn showed significant differences between the years (P<0.05). Based on the above 15 mineral elements, the origin discrimination model established with the samples of 2020 as the training set could 100% distinguish Jinghe and Zhongning wolfberry. The blind sample in 2021 was used for verification, the discriminant accuracy was 73.68%. The discrimination accuracy of the model based on the samples from 2020 to 2021 was 96.12%, and the discrimination accuracy of the model based on the samples from 2022 was 66.67%. 【Conclusion】 Although the discriminant model based on a variety of mineral elements in a single year can effectively distinguish Jinghe and Zhongning wolfberry, and the accuracy rate can reach more than 95%, the accuracy rate of discriminating blind samples is relatively low, indicating that the stability and reliability of the discriminant model has a strong interannual effect.