The volatile components in 11 cigarette samples were analyzed with SDE-GC/MS, and cluster analysis of 90 component peaks were carried out with SPSS software according to Euciolidiam distance group-average coefficient method. The results showed that the 11 cigarette samples could be grouped into 6 categories exactly corresponding to their respective brands and producers, namely, the cigarettes of the same brand and manufactured by the same factory would be well grouped into the same category. The method provides reference for brand distinguishing, quality control, comprehensive evaluation and authenticity discrimination.