Currently the recognition of psychotic illness and the recognition of improvement or recovery from such illness depend prmarily upon ascertaining the presence or absence of psychotic signs and symptoms. When we began a project (Veterans Administration Psychiatric Evaluation Project) involving evaluation of the effectiveness of the treatment of psychotic patients in 12 Veterans Adminstration neuropsychiatrie hospitals, we felt the need for techniques which would make it possible to describe the symptoms of patients in some standard form, to describe change quantitatively, and to make possible prognostic groupings of patients on the basis of their symptoms. Nothing worthy of respect can be accomplished in such a venture without scales or other devices with which different raters can secure a satisfactory level of agreement or reliability. For these purposes a psychotic symptom, no matter how interesting, is of no use as an indicator unless a number of raters can reach reasonable agreement on its absence or its presence and/or degree.