Abstract: In order to discriminate counterfeit cigarettes from genuine correctly and rapidly, cigarettes of brand A were scanned by the NIR spectrometer, an identifying model for brand A was established by factor analysis, and 88 cigarette samples of brand A, 80 cigarette samples of brand non-A and 10 samples of counterfeit cigarette were identified by the model. The results showed that the discrimination accuracy of brand A, brand non-A and counterfeit cigarette was 92.0%, 96.3% and 100%, respectively, and the average was 94.4%. It indicated that the NIRS model could be used as an auxilliary means for identifying counterfeit cigarettes.