The feasibility of UV-visible（UV-Vis） diffuse reflectance spectroscopy combined with chemometrics for the identification of Angelicae Sinensis Radix（ASR） and its adulterants was investigated in this research. A total of 40 pure and 75 adulterated ASR samples were prepared. The Kennard-Stone（KS） grouping method was firstly used to divide the dataset into a training set with 86 samples and a prediction set with 29 samples. Then the raw spectra and the preprocessed spectra by continuous wavelet transform（CWT），CWT-SG smoothing，and CWT- SG smoothing-autoscaling are analyzed. Spectra and their corresponding principal component analysis（PCA） plots were used to determine the optimal preprocessed method. Finally，partial least squares-discriminant analysis（PLS-DA） was used to build discrimination model. The results showed that PCA of the raw UV-Vis spectra could not separate ASR from their adulterated products. After the spectra preprocessed by CWT-SG smoothing-autoscaling，not only the pure ASR and its adulterants can be completely separated in the principal component space，but also the adulterants with different contents can be further identified. The optimal preprocessing method combined with PLS-DA can achieve 100% prediction accuracy for the identification of ASR with its adulterants in prediction set with one factor. Therefore，UV-Vis combined with CWT-SG smoothing-autoscaling PLS-DA is a rapid and accurate method for discriminating ASR and its adulterants.