There is still controversy surrounding the use of opioid medication for patients with chronic nonmalignant pain. Schofferman has argued that long-term opioid use leads to a "downhill spiral" associated with loss of functional capacity and a corresponding increase in depressed mood. The present study was a retrospective comparison of opioid users vs. non-users to determine whether: (a) users have higher levels of disability, medical visitation, depression, and pain; (b) the behavioral problems associated with opioid use persist after controlling for the influence of other medication; (c) opioid use is in fact a predictor of illness behavior; and (d) higher levels of opioid consumption are associated with higher levels of disability and depression. A consecutive series of 243 patients with nonmalignant pain about to enroll at a tertiary clinic were retrospectively assigned to either an Opioid User (n = 87) or Non-User (n = 156) group. Compared to Non-Users, Opioid Users were more likely to be physically disabled ( P <0.05) and depressed ( P<0.05), as well as more likely to report pain at higher levels (P<0.001) and in more locations ( P<0.05). Despite the appearance of a downhill spiral, we were unable to demonstrate an association between opioid use and any measure of illness behavior after controlling for benzodiazepine use (with the possible exception of domestic disability). Instead, we found that benzodiazepine use was significantly associated with activity level ( P<0.05), medical visitation ( P<0.01), domestic disability ( P<0.01), depression ( P <0.01), and to a lesser degree, disability days (P<0.1). Using somatization as a reference variable, we found that opioid use failed to explain a comparable amount of variance in illness behavior. Finally, there was no evidence that higher levels of opioid use were associated with higher levels of disability or depression.