Recent surveys have shown that idiopathic structural scoliosis of mild degree is generally not progressive. We will propose a mechanism which may be responsible for deterioration in the few. It has been observed that the spinal cord, although displaced towards the concavity, does not rotate in company with the vertebrae, thus exposing the emerging nerve roots to the effects of traction and possibly of entrapment. We suggest that progression occurs when the neuraxis is unable to adjust to the change in the anatomy of vertebral column. Our proposition is based upon our findings in a complete spinal column obtained from a baby with structural scoliosis. Support is provided by intercostal angiography, and by observations upon normal anatomy, the pathological anatomy of mature scoliotic spines and the anatomy of contrived scoliosis in normal spines. Although our histological and electrophysiological investigations are incomplete we can demonstrate a significant increase in degenerate cells in the dorsal root ganglia at the apex on the convex side. Lack of suitable necropsy material prevents us from confirming our observations so that our report is inevitably preliminary. We enter a plea that careful examination of the neuraxis be undertaken whenever a specimen of a scoliotic spine becomes available.