To address the Coverage Hole (CH) problem in Wireless Sensor Network (WSN) caused by nodes' random deployment or running out of energy, a novel coverage hole detecting algorithm was proposed for wireless sensor networks in this paper. The location information of sensor nodes was used to build the Voronoi diagram for the monitored area. Then the distances between a sensor node and the vertex or edges of the corresponding Voronoi cell were calculated to decide the existence of coverage holes and identify the border nodes. Simulations were conducted to evaluate the performance of the proposed algorithm within different sensing ranges and nodes density. The performance comparison with the Path Density (PD) algorithm shows that the proposed algorithm has about 10% promotion both in average detection time and average energy consumption, which is important for prolonging the network lifetime.