We evaluated eighty-three patients in whom adolescent idiopathic scoliosis had been treated with a posterior spinal arthrodesis and Harrington instrumentation extending to the second, third, fourth, or fifth lumbar vertebra. All eighty-three patients completed a questionnaire, and fifty-five patients were also examined clinically and roentgenographically at a follow-up evaluation at an average of twelve years (range, ten to sixteen years). Twelve patients had a type-I curve; twenty-six, a type-II curve; sixteen, a type-III curve; and one, a type-IV curve, according to the classification of King et al. The preoperative Cobb angle of the primary curve averaged 60 degrees and ranged from 40 to 100 degrees. The curve was an average of 35 degrees (range, 15 to 65 degrees) at the most recent follow-up evaluation. Functional assessment with use of information from the questionnaire revealed an average spine score of 81 points (range, 18 to 99 points). On the basis of the score, thirty-five patients were considered to have had an excellent result; twenty, a good result; thirteen, a fair result; and fifteen, a poor result. Sixty-three (76 per cent) of the eighty-three patients had low-back pain compared with thirty (50 per cent) of sixty individuals who served as a control group. This difference was significant (p < 0.001; chi-square test). Eighteen patients (22 per cent) needed additional spinal procedures. Fourteen patients (17 per cent) did not think that the goals of the initial operation had been accomplished.(ABSTRACT TRUNCATED AT 250 WORDS)