A multi-objective optimization approach for fibrous composite plates with curvilinear fibers is proposed in the present study. The non-dominated sorting genetic algorithm (NSGA-II) is used to obtain Pareto-optimum solutions, where the presence of a trade-off relation between the mechanical performance and curvatures of reinforcing fibers has been established. The in-plane strength around a circular hole in a finite-width plate with Tsai-Wu failure criteria or the fundamental frequency of the square plate is the mechanical objective function, while the average value of fiber curvatures is used as the conflicting objective function. Numerical results were obtained for angle-ply plates and the effectiveness of the present design approach is demonstrated by widely distributed Pareto-optimum solutions.