Recently, bearingless rotation motors have become widely popular in office automation (OA) and factory automation (FA) systems due to their simple structure, high speed of operation, and high precision in positioning. In this study, a bearingless rotation motor was designed to have a large torque and levitation force because these are generally considered as indicators of motor performance. The torque and levitation force of the bearingless rotation motor can be calculated by a commercial finite element method (FEM) program, ANSYS. To robustly design the bearingless rotation motor, first, the effective design parameters were selected; we set a table of orthogonal arrays, including the design variables and parameters. The Kriging model was applied to formulate the cost function of the regression model, which can be used to evaluate the motor performance. By minimizing the cost function, it was possible to quickly realize a robust optimum design of a bearingless rotation motor. As a result, the performance of the motor was improved with regard to the torque and levitation force.