A method is described for evaluating the progress of patients with back pain. Various symptoms were scored on analogue scales and spinal motion was measured by various techniques. These data were integrated to make subjective and objective indices respectively. Reasonable reproducibility of the measurements was obtained. This technique was used to elucidate the role of the lumbar support in surgical corsets in relief of back pain. Sufferers from back pain were randomly allocated to corsets with and without lumbar supports. There was significant improvement in those with a support compared with those without. On the other hand objective changes measured with the corset removed did not differ between the 2 groups. This study indicates that the spinal support in a lumbosacral corset makes a significant contribution towards the relief of symptoms.