The invention discloses a multi-objective ecological scheduling method and system based on a decomposition culture evolutionary algorithm, and belongs to the field of hydroelectric energy optimization, and the method comprises the steps: building a multi-objective function model by taking a water level as an independent variable and taking the maximum power generation amount, the minimum ecological flow change degree and the minimum output maximization as objectives; generating a plurality of groups of weight vectors for the multi-objective function model, initializing a neighborhood index setand an archiving set, randomly generating an initial population containing a plurality of individuals, randomly selecting parent individuals from the neighborhood index set of each individual for cross mutation to obtain filial generation individuals, comparing the parent individuals with the filial generation individuals, and reserving better individuals; and when each individual in the initialpopulation finishes the operation, finishing evolution of the initial population, and when the number of evolution times meets a preset number of times, obtaining a final archive set which is an optimal reservoir scheduling scheme. According to the method, multiple targets are evaluated, the applicability is high, and an economic and reliable hydropower station optimal scheduling scheme can be obtained.