hirty-four hips with unilateral Legg-Perthes disease which had been treated with subtrochnteric varus-derotational osteotomy and made up of follow up for more than 3 years were examined to evaluate long term effects of the surgical procedure taking opposite hips as control. The observations made were as follows. 1. The material was consisted of 30 boys and 4 girls with age ranging below 3 years by 2, 4 to 6 yesrs by 9, 7 to 9 years by 18 and over 10 years by 5 in number. 2. According to Catterall's classification, 2 cases belonged to Group II, 19 cases to Group III and 13 cases to Group IV. And as for staging, 27 cases were in necrotizing stage and 7 cases in revascularization stage. 3. The average symptom duration was 5 months ranging from 2 months to 1 year 2 months. After 10 days in average of skin traction, varus derotational osteotomy was done at subtrochanteric level of femur according to Somerville's method. 4. At follow up for more than 3 years, almost all cases showed full range of motion without pain. The exceptions were 1 case with limited abduction and 2 cases with mild pain. Twenty-eight cases had less than 1cm shortening without limping, and shortening with conceivable limp was observed in 6 cases. 5. Pre-operative and final follow-up X-rays were compared for radiological assessment by measuring femoral head migration index, spherical quotient, CE angle and Mose's femoral head sphericity. The average femoral head migration index improved for 1.92 to 1.56 and CE angle improved from 14.8°to 22°. The results were good in 21 cases, fair in 8 cases, poor in 5 cases according to Mose;s concentric circle scheme. From above obervations, it seems reasonably substantiated that subtrochanteric varus-derotational osteotomy can offer good result for children with Legg-Perthes' disease who are likely to end up badly if not treated.