OBJECTIVE The current research was designed to investigate the relationship between affective disruption and psychosis in unipolar major depressive disorder and to evaluate whether some depressed patients are particularly vulnerable to subsequent psychosis. METHODS A group of 31 psychotic and 63 nonpsychotic inpatients with unipolar depression were assessed during hospitalization and then prospectively assessed for depressive and psychotic symptoms 2.4 years after hospital discharge. RESULTS The psychotic depressed inpatients had significantly higher rates of psychosis during the posthospital period, 2.4 years later, than the originally nonpsychotic depressed inpatients. However, they did not have higher rates of depressive syndromes during the posthospital period. Psychosis during the posthospital period was significantly related to recurrent or persistent affective disruption. The posthospital depressive syndromes of the initially psychotic depressed inpatients tended to be slightly more severe than those of the initially nonpsychotic depressed inpatients. Despite this, psychosis during the inpatient period was more predictive of posthospital psychosis than the severity of the posthospital depressive syndrome. CONCLUSIONS The interaction of the following two factors was proposed to be central to psychotic depression: 1) a vulnerability to psychosis, and 2) concurrent affective disruption.