In the production-distribution system with distributed decision environment, supply chain collaborative planning was difficult to solve and reach global optimization with information asymmetry. Based on the theory and method of multi-level programming, this paper constructed a production-distribution collaborative planning model and designed a method based on fuzzy theory and genetic algorithm for solving the model. The method was an available satisfactory solution which met the constraints. Each decision-maker decided whether to accept it according to their preference and constraint conditions. If they were unsatisfied with the solutions, they could adjust the membership function to solve it again. The decision-making process had certain flexibility. At last, the validity and feasibility of the model and approach were validated with a simulation experiment.