Designs of experiments with orthogonal array do not only reduce experimental numbers but also can solve trade-off problems. This paper explains how to use this type of DoE to solve trade-off by using model experiment of rubber compound. At first, experiment should be designed by orthogonal array ( L8, L9, L12, L16, etc.) because many factors are needed to solve trade-off problem. Then, contribution % for all properties of each factors are important. The Trade-off problem will be solved if many factors can be used for the properties having higher % of contribution. Furthermore, a procedure of multiple regression analysis is explained. It is important to understand some phenomena.