This study investigates how types of bundling such as mixed-leader bundling (Buy X, get Y free) or mixed-joint bundling (Buy X and Y for $) affect the consumers' perception of the reference price and willingness to pay for the primary product. The moderating effect of product similarity and consumers' preference was also explored. Furthermore, through examining the mediating effect of the discount attribute, cognitive style, and attention level, several mechanisms were suggested in this study to better explain the effect of bundling type on consumer's price perception.The results of study 1 showed that consumers’ IRP and WTP for the primary product was significantly higher when it was offered in a mixed-leader bundling, rather than in a mixed-joint bundling. However, this bundling type effect became weaker when the bundling products were similar, rather than when they were not similar. In addition, although the moderating effect of consumers’ preference was insignificant, further analysis of the three-way interaction (Bundling Type × Product Similarity × Consumers’ Preference) showed that, in order to protect the value perception of primary product, the bundling type is most effective in a low similarity and primary product preferred situation, and is least effective in a high similarity and tie-in product preferred situation.Study 2 replicated the findings of study 1 by selecting two products that are at the same price level. Different from study 1, the moderating effect of consumers' preference was significant and consistent with the assumption. Additionally, results about the discount attribution's mediating effect showed that compared with the mixed-joint bundling, consumers tend to attribute the total bundling discount to tie-in product in the mixed-leader bundling, even when the regular price of two products is about the same. This unevenly discount attribution to tie-in product will maintain the consumers' price perception of the primary product. In the meantime, high product similarity and consumers' preference for the primary product will weak the effect of bundling type on consumers' discount attribution. At last, the bundling type effect on the price perception difference between primary and tie-in products was confirmed. Product similarity and consumers' preference also turn out to play a significant part in moderating this relationship.Study 3 further explored how product similarity affects consumers' discount attribution in the mixed-leader bundling. Results showed that low product similarity (or high product similarity) prompts consumers to use a systematic thinking style (or intuitive thinking style) and led them to pay more (or less) attention to two bundled products. In this process, the bundling discount will be unevenly (or evenly) distributed to primary and tie-in product.Based on the above results, setting the primary product in a mixed-leader bundling might be a safer choice for retailers who wish to maintain the consumers’ price perception. Furthermore, the moderating effects shown in this study provide valuable implications into how bundle products should be selected. The decision of which products to bundle is obviously important to the success of the bundling. At last, mediate variables, such as cognitive style and attention level, which suggested in this study also offer meaningful direction in designing a promotion advertisement without hurt the consumers' price perception of the primary product.