This paper addresses the control problem of a stochastic recovery system with two stocking points and different leadtimes for production and remanufacturing. For such systems the optimal control policy for a linear cost model is not known. Therefore, in the literature several heuristic policies are investigated and analyzed. In this paper a new approach is provided which differs substantially from the existing ones. Instead of using one inventory position for the production and remanufacturing decisions we base the decisions on two aggregate variables which are defined different for different leadtime relations. By means of numerical examples we illustrate that system performance, measured in average costs per time unit, can be improved substantially by our new approach, especially for large leadtime differences.