With the rapid development of the social economy, the shortage caused by the unreasonable utilization of water resources has gradually become an important factor that seriously affects normal production and life. Therefore, how to allocate water resources reasonably through scientific and effective methods has become an urgent problem to be solved. A water resource optimization allocation model with economic, social, and ecological objectives was established, and an adaptive improved Non dominated Sorting Genetic Algorithm II (NSGA-II) was proposed. Finally, it was applied to model solving. The results indicate that, the spacing values of the improved NSGA-II algorithm in the test function are 0.0054, 0.0085, and 0.0068, respectively, and the minimum mean IGD (Inverted General Distance) index is 1.177E−02. Among them, IGD is a comprehensive evaluation index that reflects the convergence and distribution of the population, while Spacing index reflects the algorithm’s ability to achieve uniformity in the optimization process, which verifies that the convergence and distribution of the method are both superior; In the allocation of water resources in Datong City, the economic benefits of the three schemes obtained are 6437.56 million yuan, 6646.97 million yuan, and 6647.36 million yuan, respectively. They can match the corresponding optimal schemes for different situations, providing new ideas for further optimization of water resource allocation and having certain technical application value.