A recent study suggests a correlation between idiopathic scoliosis in adolescence and eating disorders. However, this does not correspond with our clinical experience in the same population. The aim of this study was to verify the correlation between scoliosis and eating disorders in adolescence. A cross-sectional study was designed including 187 consecutive adolescent girls with a diagnosis of idiopathic scoliosis (mean Cobb angle 26 degrees, range 11-73 degrees, age 15.2 plus or minus 2.5; 24% juveniles, 76% adolescent type) and 93 schoolgirls as controls (age 14.9 plus or minus 1.0). All of the participants answered the Italian validated questionnaire EAT-26 about eating habits in order to identify any eating disorders. Body mass index (BMI) was calculated for all participants and compared to reference data. Statistical Analysis: chi-square test, Student's t-test, Pearson's correlation coefficient. Only 3 (1.6%; 95% CI -0.2-3.4%) participants in the scoliosis group showed EAT-26 scores suggestive for eating disorders versus 7 (7.5%; 95% CI 2.2-12.9%) in the school population (p less than 0.05). The BMI was slightly lower (p less than 0.05) for scoliosis patients (19 plus or minus 0.2) than for school girls (21 plus or minus 0.3). EAT-26 is recognized among the most valid questionnaires for eating disorders and has been widely applied in various countries. By applying this questionnaire, a lower incidence of eating disorders in female scoliosis patients was found than in the general population (using both our own controls and Italian reference values). This contrasts with some expert opinions and a recent study performed in Italy. The low BMI already reported in the literature as being typical of scoliosis participants is confirmed by our data. (Contains 1 figure.)