We reviewed the clinical records, arthrograms, and roentgenograms of sixty-one children who had seventy-two congenitally dislocated hips in an effort to identify factors that can be used to predict the outcome of treatment. Only patients who had been followed clinically and roentgenographically for a minimum of two years after the initial closed reduction were included in the study. The mean age at the time of closed reduction was thirteen months (range, three to forty-one months). The mean age at the time of the most recent follow-up was six years (range, two to fifteen years). The mean duration of follow-up after the initial closed reduction was five years (range, two to thirteen years). There was no statistical difference between the good, fair, and poor-result groups with regard to sex, the age at the initial reduction, the traction station, the side of involvement, the initial acetabular index, the initial grade of displacement, the effect of adductor tenotomy, and several of the arthrographic measurements. The medialization ratio (the percentage of the horizontal radius of the cartilaginous femoral head that lay medial to the Perkins line), measured at the time of the reduction, was significantly different (p < 0.04) between the hips for which the result was good or fair and those for which the result was poor. The medialization ratio averaged 75 per cent in the hips for which the result was good, 66 per cent in those for which the result was fair, and 57 per cent in those for which the result was poor. Limbus shapes 5 through 8 were associated with avascular necrosis (p < 0.05) and a poor result (p < 0.03).