The current physical strength test for student doesn\'t take the size of physique, which is a generic and environment factor, into consideration. The purpose of this study was, there fore, to prepare appropriate physical fitness evaluation criteria, after physical strengh was divided into three dimensions according to the extent of physical growth: good, fair and poor. The subjects in this study were 3,121 elementary schoolers who included 770 boy fifth graders, 958 girl fifth graders, 772 sixth graders and 621 girl sixth graders. They were vlassified into good, fair and poor groups according to physical condition. A multiple regression procedure was employed, by setting the physique factor as a dependent cariable and the physical strength factor as an independent one, to track the characteristice of the two factors and their interrelationship . In order to make a comparative analysis of grading between the current test and th physique-considering oe, x2 test was utilized on grade distribution. the collected data were analyzed with SAS(V.6.12) program, and the findings of this study were as follows: Firtst, among the students in their fifth year, the girl students were superior to the boy in physical condition. In the case of the sixth graders, the girl students were better than the boys in every physique factor except weight. concerning physical fitness, the boy students were more excellent than the girls in every test item, except for forehand stooping of upper body. Second, every physical strength factor was under the influence of physical condition for both of the fifth- and sixth-year groups, though its predictability wasn\'t so good. Third, the grade distribution of the current test was different from that of the physique-considering one. The former\'s distribution was irregular, but the latter tended to show a normal distribution that leaned toward the third grade. The above-mentioned findings suggested that the physical fitness grading that allowed for physical condition is a proper evaluation method that coincides with the assumption that there would be a normal dkstribution of phywical fitness among fifth and sixth elementary schoolers. Accordingly, the use of this method is recommentable.