Avascular necrosis of the femoral head has been recognized as one of the most serious and frequent complications following the initial treatment of congenital dislocation of the hip. It is now well accepted that this particular complication is iatrogenic and may be avoided by careful and adequate treatment. The reported incidence of avascular necrosis is variable because of various methods of treatment and different diagnostic criteria. We retrospectively reviewed 130 patients(144 hips) with congenital dislocation of the hip, who were treated at Seoul National University Hospital during 10 year-period from January 1974 to December 1983, and found 13 patients (13 hips) with avascular necrosis of the femoral head. We analyed avascular necrosis following treatment of congenital dislocation of the with respect to various methods of treatment, possible causes, types of avascular necrosis, to radiological findings, and to functional results. 1. The incidence of avascular necrosis following treatment of congenital dislocation of the hip in our series was 9.0%(13/144 hips). 2. The average age of the patients with avascular necrosis was 3 years, the youngest being 4 months and the oldest being 10 years. The average follow-up period was 2 year and 2 months. 3. The incidence of avascular necrosis in the closed reduction series was higher in those between 6 months to 18 months with 10.5%, as compared to 7.1% in those 6 months and under 6.7% in those 18 months and over. In the open reduction series, the incidence increased as the age increased, being 5.9%, 12.4%, 15% in the groups 18 months to 3 years, 3 to 6 years, and 6 years and over, respectively. 4. The incidence of avascular necrosis in those who had priliminary traction was 8.3%, where as that in those who had not was 16.6%, being twice as much. 5. The incidence of avascular necrosis decreased as the traction time prolonged, being 9.0% in 7 days or less, 7.7% in 8 to 14 days, 7.1% in 15 to 21 days and nil in 22 days or more. 6. The incidence of avascular necrosis in the skin traction group was 5 times as much with 11.5%, as in the skeletal traction group with 2.2%, despite older age in the latter. 7. The incidence of avascular necrosis in the open reduction group was slightly higher with 9.5%, than in the closed reduction group with 8.6%. 8. All 6 cases of avascular necrosis following closed reduction occurred when Lorenz or frog-leg cast was applied initially after reduction despite attention to avoid extreme abduction. Avascular necrosis were not encountered in those who had Lange or human position cast as the initial form of immobilization. 9. In the closed reduction, avascular necrosis occurred 8.1% in those who had adductor tenotomy and 10.0% in those who had not. In the open reduction, avascular necrosis occurred 7.9% in those who had adductor tenotomy and 18.1% in those who had not, suggesting significant role of adductor tenotomy in the prevention of avascular necrosis. 10. Avascular necrosis was attributable to excessive abduction in 7 hips, to undue pressure of femoral head in one hip, to open reduction in 3 hips, to post-operative infection in one hip, and to non-union of subtrochanteric osteotomy in one hip. 11. Of the 13 hips, following roentgenograms were available in 11. According to the classification of Bucholz and Ogden, type I, II, III, IV were 7, 0, 3, and 1 hips, respectively. 12. Functional results of the 13 hips of avascular necrosis, according to Kalamchi and MacEwen's criteria, were Good, Fair, Poor in 9, 3, and 1 hips, respectively.