Abstract: Different from existing literature on holiday economy, this study explores the effect of holidays on consumer's purchase behavior empirically based on consumer's purchase records of retailing scanner data using microeconometric methods. Using beers as the representatives of seasonal goods, category analysis indicates its sales are significantly seasonal and holidays are the demand peaks. Applying multinomial logit model to the consumer's purchase records, this study finds that price and discount are the main factors that affect consumer’s purchase decision, and that consumer’s willingness to pay for discount goes up significantly during holidays. Therefore, given the same amount of discount, more sales can be obtained in holidays than in non-holidays.