During a 10-year period, 120 drugfree DSM-III-R schizophrenic patients were consecutively and unselectively admitted to a ward for young psychotic patients and subjected to a battery of examinations including symptomatology, cerebrospiral fluid (CSF)-biochemistry, computed tomography (CT)-scan, neurophysiologic and psychophysiologic (Electrodermal activity, EDA) parameters before antipsychotic treatment was initiated. After discharge, the patients were longitudinally followed with rating of outcome (Strauss-Carpenters outcome scale) at years 1, 3, and 5 after index admission. The aim of the study was to find possible early markers for outcome in schizophrenia. At 5 years, 30% of the patients had a good outcome (total score >13) and 15% a poor outcome (total score <8). Poor premorbid adjustment and low level of education as well as negative schizophrenic symptomatology at index admission were associated with a poor outcome 5 years later. Positive symptomatology and a family history of schizophrenia did not predict outcome. Patients with a poor outcome (total score <8) had a significantly more deviant CSF HVA/5-HIAA quotient than those with a very good outcome (total score >15) as compared with healthy controls. Further, the CSF-peptides neuropeptide Y, dynorphin A, and CRF were predictable for outcome at the 5-year follow-up evaluation. Male schizophrenics who were ''nonresponders'' on the EDA test showed an almost 100% poor outcome, which was not found in females. In summary, several clinical and biological variables seem to have a predictable value for outcome in schizophrenia and, early identification of them might be a challenge for our future treatment strategies.