Fifteen subjective variables reflecting the severity of back pain, the circumstances exacerbating symptoms, and the impact of the problem on life style were scored on Visual Analogue Scales. It was possible to combine the results from these assessments to provide a global subjective index. Repeated measurements by the same observer showed a high degree of reproducibility, but when performed by separate observers discrepancies arose in certain questions. By improving certain questions and standardizing their presentation to the patient, a considerable improvement in the correlation of results between observers was obtained. Objective assessments of spinal motion and straight leg raising and a global objective index showed a high degree of intraobserver reproducibility. This technique was applied to a study of relief of back pain by lumbar corsets when it was found that there was significantly greater relief of back pain by a corset with a lumbar support than one without a spinal support. We conclude that the emphasis in assessing the progress of the back-pain patient must be on the subjective parameters, and the technique developed offers a useful method for reliably assessing patients and following their progress.