Objective: Dropout from treatment has serious implications for patients, clinicians, and researchers. The aim of this study was to examine rates of dropout from outpatient treatment for anorexia nervosa (AN) and critically examine the various definitions of dropout used.Method: A systematic review was conducted, including pharmacological and psychological interventions. All articles in PubMed, Web of Science, and the Cochrane Library were considered, and screened against a priori inclusion/exclusion criteria. Relationships between treatment outcome and dropout rate were examined across studies.Results: Nineteen relevant studies were identified, with dropout rates ranging from 4.8% (family therapy) to 100% (dietary advice). In most cases, dropout was in the range of 20-40%. Definitions of dropout used varied widely. A significant negative correlation was found between rate of dropout and body mass index (BMI) at 1 year, but this did not remain significant when differences in BMI at baseline were taken into account.Discussion: High rates of dropout from treatment for AN have serious implications for recovery, research, and the development of new treatments. A suggested reporting structure is proposed, with the aim of increasing the consistency of dropout reporting and facilitating greater understanding of this phenomenon. V V C 2011 by Wiley Periodicals, Inc.