Clinical research in patients with low-back pain is impeded because there is no quantitative measure of dysfunction. Such a measure would be useful as a control for the condition''s severity in subjects to be admitted to clinical trials, as well as a pre- and post-treatment measure to evaluate success. An approach to developing a scale to measure low back dysfunction is reported. The 105-point rating scale consists of 8 parameters grouped into 3 major parts: physical measurement of trunk strength and range of motion (40 points); patient''s perception of pain and dysfunction assessed by an activities questionnaire and a visual analogue pain scale (40 points); physician''s perception of dysfunction based on report of pain and medication usage (25 points). The rating scale was tested in 29 patients undergoing lumbar surgery and 48 patients treated in a 3-wk rehabilitation program. In these rehabilitation patients when both the patient and the physician perceived that the treatment was a success, there was a mean increase of 12.8 points. If both perceived the treatment was unsuccessful there was a mean drop of 1.4 points. The difference between these 2 subgroups is statistically significant (P = 0.011). To improve on the rating scale''s discriminatory ability, principal component analysis was performed. This analysis predicts that reweighting the 8 parameters of the rating scale will improve its performance. Utilizing a scale or index to determine treatment outcome in clinical trials should help to accurately discriminate between effective and ineffective treatment modalities.