This paper applies game theory to studing dynamic channel selection and pricing problem based on strategic customer behavior． Suppose there is a manufacturer who sells a fixed number of products to customers directly in a finite horizon with two periods． In period one，there is value uncertainty; in period two，the demand is random． In each period，the manufacturer needs to decide whether introducing a retail channel with value-added service to its existing direct channel or not，and the corresponding pricing strategies． Customers need to decide when and which channel to buy so as to maximize their own surplus． Our results show that it is better for the manufacturer to introduce a retail channel． However，the optimal pricing strategies are different when the retail channel is introduced in different periods． Meanwhile，we find that it is optimal for the manufacturer to introduce the retailchannel in the whole horizon when the retailer’s service ability is small，but not when the retailer’s service ability is very big． If the manufacturer and the retailer choose to cooperate，it is optimal to introduce the retail channel in the whole horizon regardless of the retailer’s service ability． Some results are counter-intuitive．