We analyzed 27 cases of Legg-Calve-Perthes disease that were treated by operative methods at the Department of Orthopedic Surgery, Yonsei University College of Medicine from 1975 to 1981. The modalities of treatment were innominate osteotomy in 17 patients; subtrochanteric osteotomy of femur in 4 patients; and adductor-iliopsoas tenotomies in 6 patients. Most of the cases were in Catterall group III or IV diseases with two or more of “head at risk”signs in the preoperative roentgenogram. All but one case were aged more than 7 years at the time of surgery. In all cases clinical and radiological evaluations were performed until the affected femoral head had been fully regenerated. The average period of follow-up was 2 years and 11 months. Final result by Catterall's criteria was good in 15 cases (56%), fair in 7 cases (26%) and poor in only 5 cases (18%). Among 17 cases treated by innominate osteotomy, 13 cases (76%) were in good result and there was no poor result; 15 cases (86%) showed normal containment at the last follow-up x-ray. To compare these end results with those of noncontainment treatment reported by many authors, we were encouraged to proceed our current policy of early definitive containment surgery especially in patients with poor prognostic factors. In cases of estabilshed femoral head deformities that can be confirmed by preoperative arthrography, adductor-iliopsoas tenotomies were useful to restore normal range of motion of hip joint and to prevent further subluxation of the femoral head. The significant prognostic factors at the time of surgery were the patients age; stage of disease; involvement of femoral head; and finally early established irregularities of the femoral head. In every cases of Legg-Calve-Perthes disease these factors should be carefully assessed to make proper decision for the choice of the most effective modality of treatment.