Avascular necrosis is a complication of treatment of developmental dislocation of the hip (DDH) which frequently results in significant deformity of the hip joint. Although the reported incidence of avascular necrosis varies, there is general agreement that it occurs primarily in hips that are treated and that its incidence is dependent on the age of patient at the time of reduction. The purpose of this paper was to study i)the incidence of avascular necrosis and influence of associated factors ii)the incidence of acetabular dysplasia in a series iii)to analyze acetabular development in treated hips affected by avascular necrosis and in treated hips free of this complication. From September 1983 to September 1993, 158 hips of DDH have treated in Department of Orthopaedic Surgery, Guro Hospital, Korea University. Among them, the 73 hips which had been followed-up more than 2 years after reduction, free of associated disease at diagnosis and treatment first initiated in our institute were analyzed. The following results were obtained. 1. Of the 73 hips, 27 hip (36.9%) had some evidences of avascular necrosis; based on Salter's criteria 8 hips (10.9%), Gage and Winter's criteria-19 hip (26.0%). 2. Incidence of avascular necrosis was l6.7% (2/12) at age less than 6 months, 48.6% (17/35) between 6 months to l8 months, 30.8% (4/13) between 18 months to 3 years and 30.8% (4/13) in over 3years. 3. There were no statistically significat difference of avascular necrosis between hips with closed reduction and open reduction: closed reduction - 38.6% (17/44), open rection -34.5% (10/29). 4. Cases with acetabular index (AI) of > 25degrees two or more years after reduction were considered suggestive of dysplasia. On terms of acetabular remodelling, hips with avascular necrosis showed less tendency of remodelling compared hips free of avascular necrosis; incidence of acetabular dysplasia with avascular necrosis was 70.3% (19/27), hips without avascular necrosis was 26.1% (12/46). 5. Acetabular index improved progressively after reduction, more markedly between the first 6 months and a year, then plateaued at 2 year after reduction. 6. The center head distance discrepancy (CHDD) at postreduction 1 year was reliable in terms of predictable factors for acetabular dysplasia of DDH after reduction.