We compared the clinical and radiographic outcomes between femoral varus osteotomy (23 hips) and Salter innominate osteotomy (18 hips) for treatment of Catteral is groups III and IV Perthes disease after 3-12 years follow-up. There were no statistically significant differences in the clinical outcomes using the lowa hip rating score and leg length discrepancy and in the final radiographic outcomes using the femoral head sphericity and Stulberg type between the two groups. However, neckshaft angle and center-edge angle were closer to normal value in the Salter innominate osteotomy group compared with the femoral varus osteotomy group. When the patients underwent femoral varus osteotomy at the older age (7 years), articulotrochanteric distance ratio and neck-shaft angle were significantly less than those of other patients. Salter innominate osteotomy may be better indi- cated as compared to femoral varus osteotomy, when physeal damage of proximal femur is obvious or highly suspicious particularly in the older children (7 years) with severe Perthes disease.