In this paper, we will study the calculation of the optimized weight of laminated composites under a load equals their ultimate strength. A variety of fracture criteria have been defined for this class of composite in literature; in this study we use Tsai- Hill theorem as the fracture criterion. The base of this theorem is the Classical Lamination Theory (CLT). A DOT algorithm is employed as the optimization technique with a direct use of real variables. Since optimization is a very time consuming process, we will use Radial Basis Function Neural Networks (RBFNN) to predict output. The optimization process continues in FORTRAN environment until the optimum result is achieved.