This paper extends the analyses performed in part one to the area of complex intrafamily comparisons. Ratings by patients are consistently elevated compared to ratings by parents on all scales of the Family Assessment Measure (FAM), and there are significant interaction effects, which, when examined, suggest that patient scores improve while parental scores are relatively unchanged over the course of treatment in a day hospital. The clinical implications of these findings are discussed, and we attempt to clarify the role of the family therapist in the treatment of bulimia nervosa.