In a prospective study of the course of illness of 86 schizophrenic patients (ICD 9), outcome data were obtained for 86 percent 1 year after clinic discharge. The Strauss-Carpenter Outcome Scale (frequency of social contacts, employment duration, symptomatology, and duration of rehospitalization) and the Clinical Global Impressions were used to assess outcome. The prognostic scales developed by Vaillant, Stephens, Phillips (in the abbreviated version by Harris), and Strauss-Carpenter were used as potential outcome predictors at the time of index admission. The findings were as follows: The outcome criteria used correlated at best moderately with one another. Between 60 and 90 percent of the sample were judged to have an unfavorable prognosis on the Vaillant and Stephens scales. In contrast, the outcome was relatively favorable for 50 to 60 percent of the patients. In prognostic validity, the Strauss-Carpenter scale was superior to all of the other scales investigated. However, a prognostic relationship was only established for social outcome. Social outcome dimensions showed the highest prognostic validity (e.g., employment and social contacts). The amount of variance explained by the best predictors was between 12 percent (symptomatology) and 55 percent (employment duration).