In a prospective study by the Scoliosis Research Society, 286 girls who had adolescent idiopathic scoliosis, a thoracic or thoracolumbar curve of 25 to 35 degrees, and a mean age of twelve years and seven months (range, ten to fifteen years) were followed to determine the effect of treatment with observation only (129 patients), an underarm plastic brace (111 patients), and nighttime surface electrical stimulation (forty-six patients). Thirty-nine patients were lost to follow-up, leaving 247 (86 per cent) who were followed until maturity or who were dropped from the study because of failure of the assigned treatment. The end point of failure of treatment was defined as an increase in the curve of at least 6 degrees, from the time of the first roentgenogram, on two consecutive roentgenograms. As determined with use of this end point, treatment with a brace failed in seventeen of the 111 patients; observation only, in fifty-eight of the 129 patients; and electrical stimulation, in twenty-two of the forty-six patients. According to survivorship analysis, treatment with a brace was associated with a success rate of 74 per cent (95 per cent confidence interval, 52 to 84) at four years; observation only, with a success rate of 34 per cent (95 per cent confidence interval, 16 to 49); and electrical stimulation, with a success rate of 33 per cent (95 per cent confidence interval, 12 to 60).(ABSTRACT TRUNCATED AT 250 WORDS)