OBJECTIVE Although there have been many studies of the outcome of anorexia nervosa, methodological weaknesses limit their interpretation. The authors used a case-control design to try to improve knowledge about the outcome of anorexia nervosa. METHOD All new female patients referred to an eating disorders service between Jan. 1, 1981, and Dec. 31, 1984, who had probable or definite anorexia nervosa were eligible for inclusion. Of these women, 86.4% (N = 70) were located and agreed to participate. The comparison group (N = 98) was a random community sample. All subjects were interviewed with a structured diagnostic instrument. RESULTS A minority of the patients (10%) continued to meet the criteria for anorexia nervosa a mean of 12 years after initial referral. Even among those who no longer met these criteria, relatively low body weight and cognitive features characteristic of anorexia nervosa (perfectionism and cognitive restraint) persisted. The rates of lifetime comorbid major depression, alcohol dependence, and a number of anxiety disorders were very high. CONCLUSIONS In the managed care/brief treatment era, therapeutic approaches with an excessive focus on weight gain that neglect the detection and treatment of associated psychological features and comorbidity may be inappropriate. Anorexia nervosa is a serious psychiatric disorder with substantial morbidity.