In order to efficiently utilize the energy in wireless sensor networks,a dual selection mechanism-based clustering algorithm ( DSMCA) effectively combining the voting mechanism with the time-driven one is proposed. In the voting process,a node casts a vote for each neighbor node with higher residual energy. The poll depends on the comprehensive evaluation value of the multiple attributes of neighbor nodes,and the weight coefficient of the multiple attributes is determined by means of the entropy weighting coefficient method. After the voting,each node maps its poll into a certain length of waiting time to participate in cluster head competition by using a conversion function. Moreover,a node with a higher poll produces a shorter time,thus being chosen as a cluster head prior to other nodes. Simulation results show that DSMCA balances the energy consumption among sensor nodes and effectively prolongs the lifetime of the sensor network.