This study examines the relationship between epilepsy and psychosis. It compares clinical, EEG, and neuropathologic data from a group of subjects who had both epilepsy and psychosis with similar information from another group of patients who had epilepsy but no evidence of psychotic illness. We examined, blind to clinical diagnosis, gross and microscopic material from whole-brain specimens from 10 patients diagnosed with epilepsy plus schizophrenia-like psychosis, nine subjects diagnosed with epilepsy plus “epileptic psychosis,” and 36 individuals with epilepsy (21 from an epileptic colony and 15 from the community at large) who had no history of psychosis (n = 10 + 9 + 21 + 15 = 55). We abstracted case histories without knowledge of pathologic findings. Epileptic colony patients had an earlier age at onset of seizures, while epileptic colony and epileptic psychosis patients had more frequent seizures. Epileptic individuals in the community died at a younger age than did epileptic patients in long-stay hospital care. Psychotic epileptic patients had larger cerebral ventricles, excess periventricular gliosis, and more focal cerebral damage compared with epileptic patients who had no psychotic illness. Epileptic patients with schizophrenia-like psychosis were distinguished from all other groups by a significant excess of pinpoint perivascular white-matter softenings. We found that mesial temporal sclerosis and temporal lobe epilepsy occurred with equal frequency in the psychotic and nonpsychotic groups; generalized seizures occurred more frequently in the psychotic epileptics and the epileptic colony epileptics than in the community epileptic controls.