To rapidly detect different manufacturers and batches of Wai Yong Kui Yang San, Raman spectroscopy combined with principal component analysis (PCA) - support vector machine (SVM) algorithm was used to classify and identify different manufacturers and batches of Wai Yong Kui Yang San. The results show that although the difference of Raman spectra is very weak, it can not be directly classified and identified by naked eyes, but PCA-SVM algorithm can accurately classify and identify different manufacturers and batches of Wai Yong Kui Yang San , and the accuracy rate is 100%. This method is rapid, accurate, non-destructive and simple, and has potential application value for classification, identification and quality control of different manufacturers and batches of Wai Yong Kui Yang San .