This study takes the water resources allocation system of Wuhan City Circle as the research object. Based on better understanding of water footprint, water carrying capacity, and multiple water resources requirements, a multi-level optimization model is developed to achieve the integration and balance of environmental, economic and social goals. The upper-level model focuses on minimization of pollutant emissions, the middle-level one reflects the economic benefits of the regional water system, and the lower-level one concerns about the equalization of regional water resources allocation. An improved multi-level interactive algorithm is adopted to solve this multi-level model, and the dynamic process between different decision-making goals is visually displayed with fuzzy satisfaction degree. Results indicate that the overall water resources ecology in Wuhan City Circle is satisfactory, especially in Xianning with a surplus of 4.13 hm2 per capita and Huangshi with a surplus of 1.07 hm2 per capita, whereas water carrying capacity deficits exit in Wuhan and Xiaogan. The supply and demand of water resources management system has basically reached a balance during the planning period. The economic benefits of the system are expected to reach 36.368×109 yuan, the emissions of chemical oxygen demand (COD) and ammonia nitrogen will reach 28.73×104 tons, and the earning inequality coefficient is expected to be 2.10. The fuzzy satisfaction degree of the model reaches 0.68, which reflects the interaction of the upper, middle and lower decision-making goals. The results of the multi-level programming model are then compared and analyzed with those of three traditional single-objective optimization models. These findings can provide valuable information on internal conflicts and trade-offs between environmental, economic, and social interests for the related government departments.