Abstract: In order to rapidly select the germplasm with high medical compositions as well as assessing the quality of Illicium difengpi from different regions, combined with the chemometrics software such as principal componential analysis model(PCA) and soft independent modeling of class analogy (SIMCA)), the fourier transform infrared spectroscopy(FTIR) was utilized to investigate I. difengpi from different regions. Meanwhile, extracting loading factors from PCA model were taken to analyze the differences of chemical composition among determined samples. The results were listed as follows: firstly, when the data of principal analysis model were preprocessed by standard normal variate(SNV)，the first three principal components could account for 87% variance information in fingerprint and the samples were formed into 7 different categories in principal component space. Secondly, the identification model of SIMCA could be successfully applied to predict unknown samples from different regions and the recognition rate was up to 100%. Last but not lest, the loading factor analysis demonstrated that the content of difengpin，magnolol, aromatic mixture from A, B and E was higher than other regions among identified samples. In addition, the differences of aromatic mixture of I. difengpi mainly embodied in content difference of quercetin among determined samples in different regions.