This study explores the utility of a pain IAT for the assessment of dysfunctional cognitive beliefs in chronic pain patients before and after a cognitive behaviour therapy. A patient group suffering from chronic pain (N=25) treated with a 4-week cognitive behavioural psychotherapy is compared with an untreated healthy control group (N=27) at two points in time. In addition, both groups completed a self-esteem questionnaire (Rosenberg-scale) and a self-esteem IAT. In the clinical group a questionnaire assessing self-reported pain cognitions was administered. The pain IAT was able to differentiate between chronic pain patients and healthy controls before the treatment. Most important, pain-related implicit associations could be shown to change over the course of treatment in the clinical group of chronic pain patients. Results provide first evidence for an application of the IAT in chronic pain research.