Abstract: Clustering is an effective approach for organizing a network into a connected hierarchy, load balancing, and prolonging the network lifetime. In this study, an assistant cluster head clustering algorithm based on cluster grading (ACHC-CG) in wireless sensor networks was proposed, it has three obvious features: Firstly, using sensor density to divide each cluster into one of three grades, grade coefficient is an important parameter to generate assistant cluster head. Second, it utilizes residual energy, transmitting distance and historical data to dynamic generate assistant cluster head. Last, assistant cluster head is responsible for exchanging data between neighbor cluster heads, while cluster head is responsible for receiving sensing data in cluster and processing to reduce the energy consumption of head to prolong network lifetime and prolong the interval of generating cluster head. Simulation results show that, compared with LEACH, ACHC-CG significantly balances nodes average energy consumption and prolong the network lifetime.