We present an optimization procedure that uses the Taguchi method to maximize the mean stiffness and workspace of a redundantly actuated parallel mechanism at the same time. The Taguchi method is used to separate the more influential and controllable variables from the less influential ones among kinematic parameters in workspace analysis and stiffness analysis. In the first stage of optimization, the number of experimental variables is reduced by the response analysis. Quasi-optimal kinematic parameter group is obtained in the second stage of optimization after the response analysis. As a validation of the suggested procedure, the kinematic parameters of a planar 2-DOF parallel manipulator are optimized, which optimization procedure is used to investigate the optimal kinematic parameter groups between the length of the link and the stiffness.