Adolescent idiopathic scoliosis is the most common spinal deformity affecting children, with a prevalence from mass screening programmes of 1-3%. Despite centuries of study, it remains a problem with no generally accepted theory of aetiology, and disagreement on its natural history and management. Because the deformity consists ultimately of gross left-right asymmetry, a study was undertaken to test the hypothesis that it might be a manifestation of developmental instability. Palmar dermatoglyphics in 112 normal subjects, 62 with non-scoliosis trunk asymmetry and 85 with defined adolescent idiopathic scoliosis were examined and both the absolute right-left difference and the ratio of this to the total were considered. There was increased fluctuating asymmetry of atd difference in those with any asymmetry, scoliotic or not, and increased directional asymmetry of ab and cd ridge counts only in those with pure scoliosis. This suggests that, at adolescence, developmental instability may result in a loss of symmetry in growth, and that in the presence of an increased developmental left-right gradient, this may be of sufficient severity to be classified as deformity and come to the attention of orthopaedic surgeons. This interpretation changes the focus of many previous observations on scoliosis and raises the prospect that developmental stability in humans has relevance to problems hitherto restricted to clinical practice.