Abstract: Microseismic source localization is a critical step in geological hazard monitoring and mine safety early warning. However, due to various interferences such as strong field noise, traditional algorithms are prone to falling into local optima and exhibit slow convergence, making it difficult to meet the demand for high-precision localization under complex underground conditions. To improve localization accuracy and optimization efficiency, a GA-PSO optimization algorithm for microseismic source localization is proposed, which is based on a hierarchical structure and multi-strategy adaptive mechanisms. First, a global coarse search is performed by the genetic algorithm (GA) to quickly obtain high-quality initial source locations. Then, an adaptive particle swarm optimization (PSO) with an “exploration-exploitation group” structure is introduced, incorporating an exponentially decaying inertia weight, dynamic learning factors, and an elite particle information exchange strategy to achieve fine-grained local optimization in the solution space. This hierarchical hybrid mechanism is designed to balance global exploration and local exploitation, thereby improving the convergence performance and localization stability of the algorithm. The proposed algorithm is tested on typical multi-dimensional complex functions, and the results show that it outperforms conventional PSO, GA, and optimized PSO algorithms in both solution accuracy and convergence speed. When applied to calibrated shot data from an actual mining area, the spatial localization error of the microseismic source is controlled within 15 m, and the accuracy is improved by 12.29% compared with traditional methods. The inversion results of the velocity model are more accurate, and the fitness function converges faster, demonstrating good robustness and engineering adaptability. By effectively integrating the global search capability of GA and the efficient local search strength of PSO, the proposed GA-PSO hybrid optimization algorithm significantly enhances the accuracy and stability of microseismic source localization in complex geological environments, providing a practical and optimized path for precise source localization.