The purpose of this study was to assess how accurately patients with eating disorders recall their symptoms after 6 to 12 months, to evaluate whether more recent symptoms are remembered more accurately, and to determine the extent to which the accuracy of recall impacts diagnostic classification. Seventy women who were enrolled in a longitudinal study of eating disorder symptoms were asked to recall their eating patterns, behaviors, and attitudes from 6 or 12 months earlier using semistructured interviews (Eating Disorders Examination and McKnight Longitudinal Follow-up Interview for Eating Disorders). Results indicated that correlations between the original and recalled data for frequency of objective binge eating episodes and vomiting ranged from r = .534 to .898 (average, r = .772), with lower correlations for subjective binge eating episodes (average, r = .335). Attitudes toward shape and weight were recalled more accurately at 6 months (average, r = .907) than 12 months (average, r = .620). κ Coefficients were higher for eating disorder diagnoses using broad than narrow definitions, with no differences between 6- and 12-month recall. Overall, agreement for depression recall was low but better at 6 months (κ = .423) than 12 months (κ = .296). These findings suggest that patients with eating disorders are at least moderately accurate when recalling most symptoms from 6 to 12 months earlier. Although broadly defined eating disorder diagnoses remained consistent, depression and narrower eating disorder diagnostic classifications showed more variability.