Methods of management of Legg-Calve-Perthes Disease (LCPD) are at present controversial. Some practical prognostic indices were imperative for proper management of the disease. .We followed up on twenty LCPD patients-14 who underwent varization osteotomy of proximal femur and 6 who underwent pelvic innominate osteotomy or shelf procedure-for more than 3 years and analyzed the results using the Stulberg classification. There were three requirements for the hip to be included in this study: follow-up until residual stage occurred, Catterall group III or IV and complete clinical and radiological data. The purpose of this study-was to determine the validity of radiological and clinical factors in predicting prognosis in surgical treatments. Results of the evaluation were as follows; 1.Age was found to be an important prognostic index. Older children aged > 9 years tended to do worse than younger children (P<0.05). Other prognostic indices were not found to be statistically significant. 2. When Stulberg classification IV atients were considered, there were common factors in the children who were more than 9 years old, were of Catterall group IV, and late fragmentation stage disease with more than 1.5 of the lateral subluxation index, regardless of treatment. 3. Although Catterall's classification had been widely accepted, Catterall group III or IV seemed to be difficult to be distinguished. Many cases that had been regarded as Catterall group IV were proved to have intact posteromedial portion of femoral head later in residual stage. 4. Even though well-known prognostic indices were important in anticipating the results, we considered additional unknown biomechanical factors. In some cases continuous deformation of femoral head would not be prevented even though necrotic portion of femoral head was well contained by surgical treatment.