The demand of urban-rural logistics (URL) has scattered characteristics, which result in difficulty in applying the existing network planning model of URL. A network planning model of URL was established to create a planning scheme that can meet scattered logistics demands. This model is combined with the hub-and-spoke network modeling method, including the hybrid hub-and-spoke network, single-assignment, and hub node quantity constraints. For this model, the minimum total cost was taken as the objective function, whereas the hub node location, assignment relationship, and direct path planning were taken as decision variables. Moreover, the model was solved using the tabu search algorithm. Finally, the model was applied to the network planning example of URL in Nanjing-Zhenjiang-Yangzhou, China. Results indicate that the optimal hub location and quantity, the optimal distribution path, and the most reasonable straight-through route can be determined by the hub-and-spoke network planning model of URL. In addition, the URL network shows the form of two-level hybrid hub-and-spoke with multi hubs and single distribution, in which the scattered demands can be concentrated on a few hub nodes and axes. Findings indicate that the network planning problem of URL under scattered demands can be effectively solved by the new model.