During the past 10 years, a number of inpatient programs have been designed for the treatment of chronic pain. To investigate the utility of the Rotter Internal-External Locus of Control Scale (I-E) and Wallston's and Wallston's Health Locus of Control Scale (HLC) in predicting outcome in an operantly-based, multi-modal program for the treatment of chronic pain, data were collected from 46 patients who completed a 5-week inpatient pain program. All patients completed both the I-E and HLC measures during the first few days of the program. They also kept a diary of their activity (reclining, sitting, standing) and subjective pain levels on an hourly basis throughout the program. An analysis of the results showed that patients designated as internals (lower third) on the Rotter I-E became relatively more active during the program than did externals (upper third). Internals on the HLC reported relatively less pain by the end of the program than externals, taking initial pain levels into account. These data highlight the importance of encouraging chronic pain patients to take responsibility for managing their pain condition. (Author/BL)