This research used a prospective longitudinal design to study differences in vulnerability to delusions over time in 234 subjects with schizophrenia, schizoaffec-tive disorder, or bipolar or unipolar affective disorder. Patients were assessed at three successive followups over a 7- to 8-year period. Over 60 percent of the schizophrenia and schizoaffective patients assessed experienced delusional activity at one of the three followups. Over 60 percent of the patients who initially had psychotic affective disorders also showed posthospital delusional activity. Significantly more schizophrenia patients than psychotic affective-disordered patients experienced consistent posthospital delusional activity at three successive followups. Unlike the schizophrenia subjects, affective patients showed a significant reduction in delusions after the first followup. After the initial acute psychotic episode that led to hospitalization, psychotic bipolar and unipolar affective patients showed a traitlike vulnerability to episodic delusional activity over time, but schizophrenia patients were vulnerable to more severe delusional activity and to more frequently recurring or sustained delusions. The study results question the views of several major theorists on the importance, persistence, and prognostic significance of delusions in schizophrenia.