Studies have shown that 27% to 38% of girls with adolescent idiopathic scoliosis have systemic osteopenia. The aim of this study was to investigate whether osteopenia could serve as one of the important prognostic factors in predicting curve progression.A prospective study was performed in 324 adolescent girls with adolescent idiopathic scoliosis who had a mean age of thirteen and a half years. Bone mineral density of the spine and both hips was measured at the time of the clinical diagnosis of scoliosis. All patients were followed longitudinally until skeletal maturity or until the curve had progressed > or =6 degrees . The univariate chi-square test and stepwise logistic regression were used to predict the prevalence of curve progression, and a receiver operating characteristic curve was plotted.The overall prevalence of curve progression was 50%. The prevalence of osteopenia at the spine and hips was 27.5% and 23.1%, respectively. A larger initial Cobb angle (odds ratio = 4.6), a lower Risser grade (odds ratio = 4.7), premenarchal status (odds ratio = 2.5), osteopenia in the femoral neck of the hip on the side of the concavity (odds ratio = 2.3), and a younger age at the time of diagnosis (odds ratio = 2.1) were identified as risk factors in predicting curve progression. A predictive model was established, and the area under the receiver operating characteristic curve of the model was 0.80 (p < 0.01).Osteopenia may be an important risk factor in curve progression. The measurement of bone mineral density at the time of diagnosis may serve as an additional objective measurement in predicting curve progression in adolescent idiopathic scoliosis. The bone mineral density-inclusive predictive model may be used in treatment planning for patients with adolescent idiopathic scoliosis who are at high risk of curve progression.