To address the issue of large positioning errors in the traditional DV-Hop algorithm for underwater wireless sensor networks(UWSN) in three-dimensional spaces, an improved gazelle optimization algorithm based three-dimensional positioning algorithm for UWSN(IGOADV-Hop) is proposed. Firstly, the node hops are corrected through dual communication radii, and the distance error between anchor nodes is weighted and corrected. Secondly, the Logistic mapping is introduced into the gazelle optimization algorithm to initialize the population, increasing population diversity. Then, a dynamic weight coefficient for position updating is introduced in the exploitation stage to improve the global optimization ability of node position calculation. Finally, the IGOA is used to replace the least squares method for calculating the three-dimensional coordinates of nodes, and the dynamic UWSN is constructed by adding mobile nodes and underwater noise to the network. Simulation results show that compared with the traditional DV-Hop algorithm and other swarm intelligence optimization algorithms, the proposed algorithm has higher positioning accuracy.