In order to address the problems that message-distributed task are backlogged by nodes in the inner community, active nodes are blindly selected, the message transmit disorderly and message storage redundancy in the Community-based Message Transmission Scheme in Opportunistic Social Networks (OSNCMTS). We proposed a Community-Based Message Transmission Scheme for High-Efficient in opportunistic networks (HECMTS) in this paper. In HECMTS, firstly communities are divided by EO algorithm and community matrices are distributed to nodes. Secondly the copies of message are assigned through the success rate of message and the community matrices to destination nodes. The active nodes’ information is collected through the active nodes’ back and forth in different communities. Some nodes are selected to finish message transmitting between communities. Finally, Message-Carried Nodes (MCN) transmit message according to the success rate that Forwarding Nodes (FN) carry message to destination. In addition, the MCN delete obsolete message through the success rate of message reaching destination and the time of caching message at the same time. The simulation results show that routing overhead and the average end-to-end delay in HECMTS is decreased by at least 19% and 16% compared with OSNCMTS.