Abstract: A new algorithm for solving multicriteria dynamic programming is proposed. It is obtained by extending the interactive satisfactory trade-off rate method for solving multiobjective static programming. By using some additional state variables, the mathematical model is transformed so that a dynamic programming problem is transformed into a static programming problem before iteration is performed. By doing so, the amount of computation is significantly reduced. Moreover, the relation of trade-off between criteria can be obtained easily. Using our method, the requirement to a decision maker is rather lenient in the process of man-machine interaction. For a class of general multicriteria dynamic programming problems, a solution which is satisfactory to the decision maker can be obtained quickly.