Abstract: To find out the characters of cigarette customer in different cities in China, 12 cities were selected for online investigation and stratification sampling to collect consumers'data of basic attributes and consumption behaviors. The differences in consumer personae between cities were analyzed. The crucial attributes of different groups of customers and cites were quantitatively characterized based on K-means cluster analysis and LDA(Linear Discriminant Analysis)model to assess the value types of consumers. The results showed that: 1)Customer personae in different cities differed significantly due to differences in economic development level, population density and location advantage. 2)The results of cluster analysis indicated that the target cities could be divided into three categories, the city group category Ⅰ dominated by young-to-mid-aged adults and mid-to-high-value consumer groups, the city group category Ⅱ dominated by mid-aged and mid-to-low-value consumer groups, and the city group category Ⅲ dominated by customer groups with higher value growth potential. 3)The discriminant results of LDA model suggested that urban economic and social indices and the consumption level, age and preference for cigarette products of different price categories of consumers were four crucial factors in value assessment of consumers. This method provides supports for effectively identifying the value types of consumers and making the marketing strategies of cigarette products.