Traditionally, the preliminary ship design process involves satisfying each design requirement sequentially, modifying the dimensions and repeating the process until all requirements are met. This approach neglects the interaction between the major design requirements (such as the effect of stability on deadweight), resulting in an acceptable instead of an optimal design. In order to achieve the most efficient and effective design, a multiple-objective optimization technique has been used. In the past, multiple-objective problems could be solved only in the linear domain using goal programming techniques. The proposed preliminary design optimization model involves a mix of linear and nonlinear goals and constraints and has been solved by a new method (see references [8, 9]). Five comprehensive examples are used to demonstrate the effectiveness of the method and to provide a basis for comparison with other published work. This is believed to be the first application of nonlinear goal programming in this field. The computer-based method proposed is new and makes an important contribution toward the automation of the preliminary design process.