Abstract: Study of a coordinated optimal operation of complex flood control system which consists of multi reservoirs, river channels and flood storage areas, is one of the most complex tasks in flood disaster mitigation. This paper introduces a Multi-objective Hierarchy Optimization Operation Model (MoHOOM), which sets maximum multi reservoirs system safety and minimum flood storage areas system losses as two objective functions and uses safe river channel discharge capacity as a constraint. The MoHOOM theoretical background is established by decomposition and coordination approach of large scale system theory and particle swarm optimization, the optimal solution is derived for each level derived by a third-order hierarchical decomposition and coordination structure. The MoHOOM is applied to optimal operation of complex flood control system in middle reaches of Huaihe River basin of China. The results show that the MoHOOM can efficiently calculate outflow hydrograph of multi reservoirs system and flood diversion hydrograph of flood storage areas system. Meanwhile, under the same initial calculation conditions, optimal operation reduces the safety indicator by 0.37 and 0.01 in Nianyushan and Meishan reservoir respectively; lowers the peak flow upon safe discharge capacity by 100 m³/s and 720 m³/s at Jiangjiaji and Runheji channel respectively; leads to 12.56 million Yuan losses in flood storage area, when compared to actual operation. Optimal operation reduces the safety indicator by 0.24 and 0.21 in Nianyushan and Meishan reservoir respectively; lowers the peak flow upon safe discharge capacity by 750 m³/s at Jiangjiaji channel; reduces losses by 3.42 million Yuan at flood storage area, when compared to rule-based operation. The MoHOOM is in favor of excavating multi reservoirs flood control capacities, reduces unnecessary losses in flood storage area under condition of ensuring safe discharge capacity in river channels, and ensures global optimization to coordinated operation of complex flood control system.