Recent research has suggested some efficacy for the use of hypnosis in the control of pain and distress in a pediatric population undergoing painful medical procedures. Here, we study a sample (N=23) of pediatric subjects undergoing burn-dressing changes and receiving either an imagery-based or control (social-support) treatment. Subjects' levels of distress were assessed with the Observational Scale of Behavioral Distress. Results indicated that distress behaviors in this population can be measured reliably using this scale. However, no support was found for the main hypothesis that imagery treatment would be superior to control treatment in the alleviation of distress, nor were these treatments effective in comparison to baseline conditions. We discuss the formidable problem that burn and dressing-change pain presents, as well as the reasons why this treatment attempt might have failed to have the predicted effects. We also discuss important developmental considerations regarding the adequate assessment of pain and distress.