Abstract: In the reserch, a method for rapid identification of Xishui brown goat by rapid evaporative ionization mass spectrometry (REIMS) was established. This study obtained a real-time traceability detection method for Xishui brown goat conducting optimization of instrument parameters (The voltage was set to 40 V, the length was 2 cm, the incision depth was 0.5 cm, and the cutting time was 2 to 3 s) and model building after collection of lamb samples from different origins. Models were established using principal component analysis and partial least squares discriminant analysis, and marker compounds with variable weight value (VIP) >1 and ANOVA non-parametric test P value ≤0.05 were selected to establish identification component score plots, Loadings plots and abundance plots to identify the characteristic markers with large differences. The results showed that the model had good clustering separation effect and could be distinguished from other groupings with the identification accuracy of 97.32%. A number of characterized compounds with large difference were identified by LIPID MAPS database including sphingosine, phosphatidylinositol, heptadecadienoic acid, eicosadienoic acid and succinic acid. The developed testing method would provide a new identification idea for the origin traceability of Xishui brown goat with various advantages of pre-treatment free, short detection time, high recognition rate and stability.