Although it has often been suggested that chronic pain patients who are receiving workmen's compensation or who have litigation pending are less likely to benefit from treatment, the results of outcome studies of this question conducted by various pain clinics have been inconsistent. We hypothesized that poorer outcome in such patients may be related to the fact that they are less likely to be working and that the inconsistent results in the literature may therefore be explained by variability among studies in the percentages of patients who are receiving compensation (or who have litigation pending) who are also working. We examined the relationships among compensation, litigation, employment, and short- and long-term treatment response in a series of 454 chronic pain patients. Compensation benefits and employment status both predicted poorer short-term outcome in univariate analyses; however, when employment and compensation were jointly used to predict outcome in multiple regression analyses, only employment was significant. In additional analyses, only employment significantly predicted long-term outcome, whereas compensation and litigation did not. Our results suggest that it would be valuable to redirect attention away from the deleterious effects of the 'compensation neurosis' and toward the roles of activity and employment in the treatment and rehabilitation of chronic pain patients.