Objective Binge eating (BE) is diagnostically important for eating disorders (EDs) but difficult to measure validly and reliably. The study compared interview and questionnaire formats regarding the proportion of patients in each diagnostic group who reported BE. Method Data came from the clinical Stepwise ED database (N = 3508 adults and 1354 children/adolescents), representing the full range of ED diagnoses. Methods used were the Structured ED Interview and the ED Examination Questionnaire. Results Opposite discrepancies were observed between methods depending on ED diagnosis, such that a higher number of restrictive patients reported BE on the questionnaire than the interview, whereas the opposite was true for bulimic patients. Discussion The findings have important clinical and theoretical implications and suggest that BE assessment is more complex than previously believed. Copyright © 2014 John Wiley & Sons, Ltd and Eating Disorders Association.