We address the discrete lot sizing and scheduling problem with returns options, assuming that only manufacturing or remanufacturing can be carried out in a certain period, but not both, and at full capacity. We relax the common assumption of a single production speed by allowing different values within a finite set of pre-established alternatives. A mixed-integer linear programming model is developed for the problem and a computational study is conducted to analyze the impact of allowing multiple manufacturing and remanufacturing speeds, considering different scenarios for the setup and unit processing cost values as well as for the quantities of returns. Among the findings of the computational study, we highlight the cost savings achieved by enabling multiple speeds, which are primarily due to less new product manufacturing (and more remanufacturing), but also to a reduction in inventory levels by better matching production quantities to demand requirements. We conclude the paper with some remarks and guidelines for expanding this research in the future.