By leveraging the 5G enabled ve-hicular ad hoc network (5G-VANET), it is widely recognized that connected vehicles have the potentials to improve road safety, transportation intelligence and provide in-ve-hicle entertainment experience. However, many enabling applications in 5G-VANET rely on the efficient content sharing among mobile vehicles, which is a very challenging issue due to the extremely large data vol-ume, rapid topology change, and unbalanced traffic. In this paper, we investigate content prefetching and distribution in 5G-VANET. We first introduce an edge computing based hierarchical architecture for efficient distribu-tion of large-volume vehicular data. We then propose a multi-place multi-factor prefetching scheme to meet the rapid topology change and unbalanced traffic. The content requests of ve-hicles can be served by neighbors, which can improve the sharing efficiency and alleviate the burden of networks. Furthermore, we use a graph theory based approach to solve the content distribution by transforming it into a maximum weighted independent set problem. Finally, the proposed scheme is evaluated with a greedy transmission strategy to demonstrate its efficiency.