Location is one of the basic properties for an object. With the development of wireless sensor network (WSN), the requirements for sensor nodes’ location have become more and more important in practical applications. Ultra-wideband (UWB) and inertial measurement units (IMU) have been widely used in WSN due to their high positioning accuracy. UWB is of high accuracy but it is susceptible to multipath effects and relative geometric position relationships between nodes. IMU can provide continuous inertial information, but cannot solve the cumulative error problem. Thus, the fusion positioning method based on UWB and IMU can compensate the UWB multipath effect and IMU error accumulation problem, and finally improve the positioning accuracy. In this paper, a new fusion positioning method based on UWB and IMU is proposed to realize the high-precision position tracking of the target nodes in sensor network. The CRLB (Cramer-Rao lower bound) is calculated to characterize the spatial location performance of the fusion positioning method, and the PCRLB (posterior Cramer-Rao lower bound) is calculated to characterize the temporal positioning performance of the fusion localization method. Both CRLB and PCRLB are used to provide theoretical support for the design and simulation of the fusion positioning algorithms. Experimental results show that the proposed fusion method has better positioning performance in both temporal and spatial aspects, which is closer to the theoretical lower bound of practical application.