Heath and coworkers proposed that schizophrenia may be an autoimmune disorder in which antibodies are built up against specific substances in certain brain cells. Heath reports that schizophrenic patients exhibit abnormal brain waves in recordings from the caudate nucleus and septal area. These abnormal waves can also be recorded from similar sites in monkey brains after injections into the lateral ventricle cerebrospinal fluid of gamma-G-immunoglobulins (IgG) isolated from the blood of acutely ill schizophrenic patients. We prepared IgG fractions from control subjects and acutely ill schizophrenic patients and tested them in rhesus monkeys under double-blind conditions. Of 107 sera tested from 24 schizophrenic patients, 29 produced positive electroencephalographic recordings in the monkeys. From 30 control subjects we tested 80 samples and found 6 to be positive according to Heath's criteria. This amounts to more than 1 positive reaction for every 4 schizophrenic patients' fractions tested and approximately 1 positive in 13 from control subjects' serum fractions. The difference between control and patient groups is highly significant (p less than 0.001). Although our results confirm the experimental findings of the Heath group concerning abnormal EEG activity associated with an IgG fraction from schizophrenic patients, they differ from Heath's results for fractions from control persons. We found positive effects from a small number of control fractions whereas Heath claims never to have observed positive biological activity in control fractions. The autoimmune hypothesis has numerous drawbacks, the greatest of which is the inability to demonstrate the presence of circulating antibody in schizophrenic patients with the use of standard immunologic techniques.