Abstract: Due to the consistency of appearance and processing, the place of origin of Wuyi rock tea can not be identified by its appearance characteristics. The elemental composition of tea leaves is different due to the different environment of producing area, forming the unique elemental fingerprint of tea producing area. By measuring the contents of various elements in different producing areas of Wuyi rock tea, the origin discrimination model was established by means of stoichiometric statistical method, which provided technical reference for the origin discrimination of Wuyi rock tea. A total of 55 tea green samples were collected from three main producing areas of Wuyi rock tea. The contents of mineral elements were determined by inductively coupled plasma optical emission spectrometry (ICP-OES) and inductively coupled plasma mass spectrometry (ICP-MS), and the data were analyzed by using the principal component analysis. At the same time, two discriminant methods, including the orthogonal partial least squares discriminant analysis (OPLS-DA) and Fisher linear discriminant analysis (FLDA), were used to discriminate the producing areas of Wuyi rock tea. The results showed that: the elements in the tea samples from the three main producing areas of Wuyi rock tea showed different characteristics. The content of the elements in the samples was tested by Bartlett’s sphericity test. The Kaiser Meyer-Olkin (KMO) value was 0.650, and KMO > 0.5 indicated that the grouping clustering effect was better. Through the principal component analysis, five principal components were successfully extracted, and their eigenvalues were more than 1, and the cumulative variance contribution rate was 71.560%. The cumulative interpretation ability parameters of the orthogonal partial least squares discriminant model were R²X = 0.652, R²Y = 0.752, and the predictive ability parameter Q² was 0.590, all exceeding 0.5. The correct discrimination rate and cross-validation rate of the Fisher linear discriminant model were more than 90.9%. Both models could accurately discriminate the tea green samples collected from three different producing areas. This study demonstrated that the mineral elemental fingerprints of Wuyi rock tea samples from three different producing areas, combined with stoichiometric statistical methods, could effectively identify their origin, providing a theoretical basis for the protection of place of origin of Wuyi rock tea.