Consistent predictors, and more especially moderators, of response to psychological treatments for eating disorders have not been identified. The present exploratory study examined predictors and moderators of outcome in adult patients who took part in a randomised clinical trial comparing two leading treatments for these disorders, enhanced cognitive behavioural therapy (CBT-E) and interpersonal psychotherapy (IPT). Four potentially important findings emerged. Firstly, patients with a longer duration of disorder were less likely to benefit from either treatment. Second, across the two treatments the presence, at baseline, of higher levels of over-evaluation of the importance of shape predicted a less good treatment outcome. Third DSM-IV diagnosis did not predict treatment outcome. Fourth, with the exception of patients with baseline low self-esteem who achieved a better outcome with CBT-E, it was generally not possible to identify a subgroup of patients who would differentially benefit from one or other treatment.