Low field nuclear magnetic resonance (NMR) technique was used in this study to detect adulterated milk (mixed with water, salt, urea and sucrose) and pure milk, combining with principal component analysis method (PCA), partial least squares discriminant analysis (PLS-DA) and linear discriminant analysis (LDA)The results showed that adulterated milks showed a certain regularity distribution with adulterant proportion in the PCA score plot, and was easy to identify. PLS-DA and LDA were used to establish the discriminant models for all adulterated milks. The accuracy discrimination rates using these two methods to detect milk adulterated with water and urea both were 100%, However, when the two methods were utilizied to investigate the milk mixted with salt or, sucrose respectively, the accuracy discrimination rates were 83.33% and 100%, or 73.33% and 76.67%. Thus, PCA, PLS-DA and LDA could be used for the rapidly analysing data of low field NMR, and LDA showed the highest identification accuracy.