Many of the individual biological, psychological, and social factors attributed to low-back disability have been tested previously for their ability to predict treatment outcomes. To test the assumption that disability exaggeration affects treatment outcomes, models were developed to quantify this complex characteristic and to test its predictive value. Two hundred fifty-eight patients with chronic back disability entering a program of functional restoration were initially evaluated with a battery of tests, including measurements of trunk flexibility, lifting capacity, cycling endurance, self-assessments of pain and disability, and psychological attributes. On the basis of these measurements, patients were characterized as disability exaggerators if by peer comparison their self-assessments of pain and disability were in the most severe range despite high levels of physical capacity. Program completion and work status 1 and 2 years after treatment were compared between disability exaggerators and their peers. Individual initial attributes associated with program completion included pain intensity and Million Visual Analogue scores, lifting capacity, trunk flexibility, some Minnesota Multiphasic Personality Inventory and Million Behavioral Health Inventory scales, and cigarette smoking. One-year re-employment was associated with Minnesota Multiphasic Personality Inventory Scale 8, Wechsler Adult Intelligence Score-Revised, and cycling endurance. There were no significant associations between any individual factor and 2-year work status. Only two of the 12 disability exaggeration models distinguished between program graduates and dropouts, and none of the models accurately predicted return to work following treatment. Prescription of intensive multidisciplinary treatment should not be denied on the basis of any individual patient attribute or of disability exaggeration, as measured in this study. Measurement of disability exaggeration may be improved by the use of taskspecific comparisons of self-perceived and observed physical capacities. Disability exaggeration may be more useful in predicting outcomes of treatments that neglect its psychosocial components.