A recent revision of the proposed DSM III description of the borderline category is presented with reference to the work of Masterson and Grinker. The contributions of Roger Shapiro and his group to a psychoanalytic formulation of the developmental and familial dynamics are summarized with reference to genetic factors and the lack of validation, so far, for any general hypothesis. Treatment recommendations of Wolberg, Shapiro and Edith Jacobson are described. Grinker's operational criteria were chosen for the pilot study: a defect in affectional relationships with no enduring close ties; child-like angry relationships with parental or authority figures, particularly in the dysphoric group; low self-esteem, very poor sense of identity; a chronic depression characterized more by loneliness and boredom than by guilt or loss, and a pervading sense of anger or resentment. Cases are presented to illustrate treatment strategies. A brief differential diagnostic statement is made describing "borderline" as a separate category from other disorders manifesting oppositional, antisocial and violent behavior. (Author)