Serum pharmacochemistry of traditional Chinese medicine (TCM) is a direct and effective method for determining efficacious substance foundation of TCM. However, the complexity of chemical constituents of TCM and the interaction among ingredients in the in vitro process make the analysis on in vitro ingredients of TCM arduous. Liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS) has become the cornerstone in detection and characterization of in vivo ingredients of TCM because of its sensitivity and ability to analyze complex mixtures. However, due to significant interference from endogenous species, detection and identification of the constituents of TCM in the biological matrices are often difficult. There is a crying need for introducing specialized ingredient identification techniques to avoid artificial omission of in vivo ingredients of TCM. On the basis of the analysis on transitional ingredients in rat blood, this essay introduces the application of such pattern recognition methods as mass defect filter, Metabolynx software and principal component analysis, partial least squared discriminant analysis and orthogonal partial least squared discriminant analysis in identifying in vivo ingredients of TCM.