This study evaluates the validity and the reliability of a new instrument developed to assess the psychotic spectrum: the Structured Clinical Interview for the Psychotic Spectrum (SCI-PSY). The instrument is based on a spectrum model that emphasizes soft signs, low-grade symptoms, subthreshold syndromes, as well as temperamental and personality traits comprising the clinical and subsyndromal psychotic manifestations. The items of the interview include, in addition to a subset of the DSM-IV criteria for psychotic syndromes, a number of features derived from clinical experience and from a review of the phenomenological descriptions of psychoses. Study participants were enrolled at 11 Italian Departments of Psychiatry located at 9 sites and included 77 consecutive patients with schizophrenia or schizoaffective disorder, 66 with borderline personality disorder, 59 with psychotic mood disorders, 98 with non-psychotic mood disorders and 57 with panic disorder. A comparison group of 102 unselected controls was enrolled at the same sites. The SCI-PSY significantly discriminated subjects with any psychiatric diagnosis from controls and subjects with from those without psychotic disorders. The hypothesized structure of the instrument was confirmed empirically.