The search for new superhard materials is of great importance in view of their major roles played for the fundamental science and the industrial applications. Recent experimental synthesis has made several great successes, but the difficulties associated with synthesis in general remain. Materials design technique is greatly desirable as a request to assist experiment. In this paper, two rational theoretical methods of design of superhard materials have been reviewed: (i) substitutional method, which is successful in some cases, but limited to the known chemically related phases, and (ii) global free energy minimization method, which can be applied to large scale of materials with the only information of chemical compositions. The successful applications have been described and the main principles are summarized.