Abstract: In order to explore the reliable method of milk powder origin traceability, 12 kinds of milk powder came from 3 regions were regarded as object of study. The contents of 7 chemical elements of Ca, K, Mg, Na, Al, Fe, Sr were detected by ICP-AES. The contents of 15 chemical elements of Li, Cr, Mn, Co, Ni, Cu, Zn, Se, Ba, Pb, Sc, Y, La, Ce, Nd were detected by ICP-MS. The content of P was detected by spectrophotometric method. Then SPSS was used to analyze the data by analysis of variance (ANOVA) , principal component analysis (PCA) and cluster analysis (CA) . Results showed that the element contents were different in the milk powder samples from different locals. PCA and CA classified the samples of milk powder into different categories, which was consistent to the variety and the geographical origin. The overall accuracy of CA achieved 91.67%.