We address the dynamic lot sizing problem for systems with product returns. The demand and return amounts are deterministic over the finite planning horizon. Demands can be satisfied by manufactured new items, but also by remanufactured returned items. The objective is to determine those lot sizes for manufacturing and remanufacturing that minimize the total cost composed of holding cost for returns and (re) manufactured products and set-up costs. Two different set-up cost schemes are considered: there is either a joint set-up cost for manufacturing and remanufacturing (single production line) or separate set-up costs (dedicated production lines). For the joint set-up cost case, we present an exact, polynomial-time dynamic programming algorithm. For both cases, we suggest modi. cations of the well-known Silver Meal (SM), Least Unit Cost (LUC) and Part Period Balancing (PPB) heuristics. An extensive numerical study reveals a number of insights. The key ones are that, under both set-up cost schemes: (1) the SM and LUC heuristics perform much better than PPB, (2) increased variation in the demand amounts can lead to reduced cost, showing that predictability is more important than variation, and (3) periods with more returns than demand should, if possible, be avoided by 'matching' demand and return.