A age-related morphological changes in bow legs and knock knees were studied on roentgenograms taken in a standing , notably on alignment of the lower extremities. In the present study comparisons of chronological changes of this alignment, which was assessed by the femoro-tibial angle, were made between 3 groups of children under 16 years of age: 350 controls (656 legs), 148 with genu varum (296 legs) and 176 with genu valgum (252 legs). Bone growth at the lower end of the femur and at the upper end of the tibia was found to influence most closely on the changes in the femoro-tibial angle. The femoro-tibial angle of the controls under 3 years of age showed more than 180 degrees, the lower extremities were bowed outwardly, and the axes of the femoral shafts were medially tilted. From the age of 3 to 5 years the lower extremities reversely tended to be knocked inwardly owing to gradual growth of the lateral condyles of the femur and the tibia. Then, with advancing age, the axis of the femoral shaft again tended to tilt outwardly and at the same time the tibial axis tended to be tilted inwardly. Since these changes in the direction of bone axes are counter-balanced, the femoro-tibial angle tended to be 177.37 degrees, approximating to the value of the japanese adults (176 degrees - 178 degrees). Then, the femoro-tibial angle can be estimated by measuring an angle between the proximal epiphyseal line and the axis of the tibial shaft. In children with knock knees there occurred no marked chronological changes in bone alignment and the lower extremities were always knocked inwardly as compared with the controls. In children with bow legs, on the other hand, the deformity tended to be normalized notably with the age advancing as a result of growth of the medial condyle of the femur. The mechanical axis of the lower extremity passed through somewhat lateral side to the center of the knee in the most of the controls aged from 2 to 5 years and through a little medial side in those of the other age group. In children with knock knees and in those with bow legs, on the other hand, the axis was always found to pass through the laternal and the medial side, respectively, to the center of the knee. The rate of deviation of the axis was noted in correlation with the femoro-tibial angle with coefficient index of 0.855 in the controls, 0.875 in a group of knock knees and 0.951 in a group of bow legs.