On the basis of principal-agent theory, strategy of quality control on web trading platform under double moral hazard condition is discussed. When the online seller's costs which are invested in products' quality control and web trading platform's costs which are invested in products' quality evaluation both are unobservable information, the moral hamd from online seller and web trading platform will be occurred. In order to reduce the double moral hazard, incentive measures are given: one is the penalty to online seller when they provide defective products and what they have done isn't being found by web trading platform, and the other is the penalty to online seller when they provide defective products and what they have done is being found by web trading platform. The quality control model under double moral hazard condition is given after the online seller's revenue and web trading platform's revenue are considered, respectively. And the simulation is done by use of the evolutionary programming algorithm.