Abstract: Based on the characteristic of high uncertainty in remanufacturing reverse logistics networks,a stochastic location model for mixed integer nonlinear programming(MINLP) is built,in which return centers and remanufacturing factories are seen as stochastic queuing systems that are of characteristics of M/M/l and M/M/c respectively.Cycle time and inventory costs of the old products in the system are taken into account.Then a bi-level genetic algorithm is proposed.The feasible combinations of 0-1 binary variables are searched by the outer genetic algorithm and the remaining transportation sub-problems are solved by the inner genetic algorithm.Finally,an example is given to prove the validity of the model and algorithm.