In order to classify the brands of Chinese liquors effectively, the fluorescence characters of Chinese liquors are compared and analyzed. The concentration scores of training samples and testing samples are obtained by using parallel factor method (PARAFAC) combined with genetic algorithm (GA). Support vector machine (SVM) method is adopted to establish the identification model of Chinese liquors, and the accuracy rate of prediction is 97.5%. The experimental results show that the concentration scores of the three principal components reflect the difference between brands. The combination of PARAFAC and GA provides an accurate method for the rapid identification of unknown samples. The results indicate that PARAFAC-GA-SVM has higher prediction accuracy. The proposed method can effectively extract the spectral characteristics, and also reduce the dimension number of the input variables of SVM. The results can provide a new way for the identification of Chinese liquors.