A method to express the flexible cam curve with continuous jerk by using quintic NURBS curve was presented. Then the cam curve could be optimized by adjusting knot vector, control points and weight factors of NURBS curve. The multiobjective optimization for a cam curve was described to satisfy both kinetic and dynamic requirements. To solve the optimization model for NURBS cam curve, a hybrid genetic algorithm combining elitism strategy, selection strategy based on goal attainment method and fast non-dominated sorting algorithm was proposed. The optimization example illustrated that the presented hybrid genetic algorithm could solve multiobjective optimization problem and achieve cam curve with good synthetic performances.