Abstract: A new fast chemical pattern recognition method was developed to discriminate traditional Chinese medicine(TCM) of Caulis spatholobi and adulterant Caulis sargentodoxae based on Fourier transform infrared spectroscopy(FTIR).The original data matrix of FTIR were pretreated with normalization and wavelet transform,and discrimination results of two kinds of TCM using principal component analysis and cluster analysis were compared before and after data processing.Results showed that two kinds of TCM could be completely and correctly discriminated using wavelet detail coefficient as the characteristic variables.Kinds of TCM in the verification set was predicted using partial least squares(PLS) calibration model with 100% accuracy of prediction.