Abstract: This paper develops a fuzzy recognition dynamic programming model for hybrid multiobjective multistage decision making problems arising from complex systems A fuzzy aggregation operator is first de rived from a fuzzy pattern recognition model to get a global evaluation for all objectives Then,by incorporating the operator into the traditional dynamic programming,two dynamic approaches are proposed to solve the problems,i e,the minimum ranking approach and the stage recognition approach,which form the fuzzy recognition dynamic programming model The two approaches are of characterist ics of different back re currence equations A nonstructural decision analysis method is simultaneously used to compute the objective weights in the dynamic process of decision makings Finally,a case study is given to show the model is rea sonable and feasible in practice.