Abstract: In this paper, the centralized decision-making model, the decentralized decision-making model of the non-Reward-Penalty Mechanism as well as the closed-loop supply chain models, which considers competitions of manufactore and is based on the Reward-Penalty Mechanism, have been established. By solving and comparing the results of these models with the method of backward induction respectively, the buyback price, wholesale price, retail price and collection decisions of closed-loop supply chain have been obtained. Studies have shown that: compared with the decentralized decision-making under the non-Reward-Penalty Mechanism, the government's Reward-Penalty Mechanism not only can effectively guide manufacturers to reduce the price of new products, improve the collection rate and improve the efficiency of the closed-loop supply chain but also is conducive to the increasing of the profits of the closed-loop supply chain members; Reward-Penalty Mechanism make the recycling manufacturer's profits increased, and the non-recycling manufacturer's profits decreased, which can guide the manufacturer's enthusiasm of collection and remanufacturing effectively; meanwhile, the increase of the recyclers' and the retailers' profits can also arouse their enthusiasm, and then make the whole closed-loop supply chain members' harmonious operation.

Key words: manufacturer-compete, closed-loop supply chain, collection rate, the Reward-Penalty Mechnanism