Ninety patients with severe anorexia nervosa fulfilling DSM-III-R criteria were assessed in depth in terms of their family developmental psychopathology and then randomly allocated to either one of three treatment groups or to no treatment. In three treatment regimes, a behavioural approach to diet and weight gain was coupled with individual and family psychotherapy directed at the adolescent maturational problems. All three treatment regimes were highly significantly effective at one year in terms of weight gain, return of menstruation, and aspects of social and sexual adjustment. Body weights above those at pubertal onset were achieved for the group mean maximum and one-year follow-up weights for all three treatment groups but not the control group.

We describe the design of a random allocation prospective treatment study comparing three modalities of care and a nontreatment control. Problems encountered in the execution of this study are discussed.

No abstract