Abstract: To address the issue of insufficient localization accuracy in traditional DV-Hop algorithms for three-dimensional positioning in wireless sensor networks, this paper proposes a three-dimensional DV-Hop localization method based on an enhanced gazelle optimization algorithm. At the algorithmic design level, the following improvements are made to the gazelle optimization algorithm: a Cubic chaotic map is employed to initialize the population, thereby enhancing the diversity of initial solutions; and the golden sine strategy is introduced to optimize the exploration phase, achieving a better balance between global search and local exploitation. In terms of localization process reconfiguration, a multi-level communication strategy is designed to refine hop counts between nodes. Optimal beacon nodes are selected based on the three-dimensional spatial distribution of sensor nodes to calculate the average hop distance. Moreover, the maximum likelihood estimation method is replaced by the improved gazelle optimization algorithm to determine the optimal coordinates of unknown nodes. Experimental results demonstrate that, within the WSN monitoring area, the proposed algorithm reduces the average localization error by 18.7% and 7.3%, respectively, compared with the traditional 3D DV-Hop algorithm and the SSA-3D DV-Hop algorithm. This method provides a new approach for high-precision node localization in complex environments and can be extended to applications such as precision agriculture and disaster early warning.