In reply

We welcome the opportunity to clarify the interpretation of our study of CBT for bulimia nervosa. This study was designed to identify those patients who responded especially well to CBT (complete abstinence from bingeing and purging) but who might benefit from additional follow-up support to consolidate their improvement. To do this, we employed a highly sensitive definition of "relapse" to mean no episode of bingeing or purging whatsoever in the month prior to the 17-week or 4-month posttreatment interview.

As Fairburn and Cooper point out, however, our article could be misinterpreted as a study of clinical relapse, thereby casting doubt on the value of CBT. This is not the case, and the following data should make this clear: