The purpose of this study was to construct a conventional motor achievement test battery based on pass or fail criteria for preschool children. Thirty seven items were examined on reliability and objectivity of both observed value (0V) and measured value (MV); agreement between OV and MV, and discriminating power of items. The thirty seven items were selected from 223 items by taking into account various conditions like a hierarchical structure in fundamental motor ability proposed by Gallahue, possibility of behavior observation, difficulty of task, and so on. As a result, twenty eight items representing each motor pattern of locomotion,manipulation, and stability domains, were reselected. For the purpose mentioned above, using factor analytic procedures, eight motor pattern factors for three domains were extracted and interpreted as folow;: 1) walking-jumping, 2)hopping, 3) climbing for the locomotion type, 1) catching, 2) throwing further, 3) throwing to aim for the manipulation type, 1) maintaining posture, 2) stability during action for the stability type. Examining the factorial validity in addition to taking practicability into consideration, 16 items were chosen at first and 9 items were chosen from them next. Because these test batteries showed a positive relation with another motor test on which performances were measured by CGS scales, they were considered to be practical and effective to estimate motor ability. Finally,assessment scales for two test batteries were constructed, using data of 3564 preschool children aged from three to six.