Abstract: An optimization model for multi-objective inventory and distribution strategies was established to solve the stochastic demand inventory routing problem with time windows(IRPTW).The model was solved with a multi-objective genetic algorithm(GA).The algorithm uses such genetic operators as best choice and adaptive strategy to approach the global optimal solution.It overcomes the inability of conventional GA in local search,increases convergence speed,and improves global optimization performance.A product distribution system of a logistic company was taken as an example,and the result shown that an optimal scheme with a reasonably low cost was obtained with the algorithm.