THERE is little disagreement among mental health professionals that current procedures for psychiatric diagnosis are largely unreliable and, hence, of doubtful validity and utility. Several sources of low diagnostic reliability have been traced.1-3Various limitations on diagnosis as an aid to prognosis, treatment, and determination of etiology have also been reviewed.4,5Because of its carefully documented shortcomings, some investigators have argued for the abolition of the process of formal psychiatric diagnosis, in favor of the "single-case" approach which views each individual's Psychopathology as unique and unclassifiable.6Others would retain the concept of diagnosis but radically alter traditional procedures so that behavior is categorized as one set or another of characteristic maladaptive interactions with the environment rather than as a collection of psycopathological symptoms.7A smaller group of researchers argue for less fundamental changes in current diagnostic procedures. For the most part, they