Online recycling has been recognized as an efficient method for waste recycling. This paper focuses on the information asymmetry between an internet recycler and consumers in the online transaction of used products. This paper is to find an optimal strategy for the internet recycler when the consumers would make an adverse selection in submitting the classification results (the used products would be classified into two kinds according to the quality: High quality and Low quality) of used products in online orders to avoid the loss because of internet recycler's moral hazard, which might bring the extra cost for internet recycler. Therefore, this study used game theory to establish a Stackelberg game model for analyzing an internet recycler and consumers' decision-making in the online transaction of used products. Based on the analysis of consumers' behaviors in an online transaction, internet recycler's strategies are divided into two kinds: A, high moral hazard strategy, and B, low moral hazard strategy. It is found that the strategy of low moral hazard is optimal for the internet recycler compared to the strategy of high moral hazard. Further, though strategy B is optimal, the internet recyclers is suggested to increase their moral hazard probability when the H used products are increasing (High-quality ones). Besides, for strategy B, the correction cost for wrong H orders and correction benefit from correction of wrong L orders would decrease the optimal moral hazard probability, and the impact of the correction benefit from correction of wrong L orders on the decision of moral hazard probability is more obvious.