Abstract: Nine tobacco flavor samples were detected by an electronic nose, the obtained data were analyzed with principal component analysis (PCA) and discriminant factor analysis (DFA). The models of quality control were established, and the quality consistency of the samples was evaluated. The results showed that: 1) tobacco flavors of different flavor notes or of certain similarity could be distinguished by electronic nose; 2) the abnormal product could be successfully discriminated from qualified products by electronic nose trained with qualified products along with established quality control model.