In practical decision making process, people often have to deal with indeterminacy data. To handle this situation, many optimization models involved uncertainty are widely discussed. Specifically, in a supply chain, costs of transportation and inventory are important factors in optimizing profits. While, the market demands are often unknown, especially when a new situation arises. In this paper, based on uncertainty theory, a new type of two-stage programming, named uncertain programming with recourse (UPR) is first put forward. Then, by employing the expected value of uncertain variable, an equivalent classic programming of UPR is built. Finally, by regarding the market demands as uncertain variables, UPR model is used to solve the integrating transportation and inventory problem under uncertainty.