Patients who fail to meet criteria for Briquet's syndrome (or somatization disorder) despite a chart diagnosis of hysteria have been shown previously to resemble patients with primary depression in terms of familial psychopathology. The same patients are shown here to have excess mortality which also resembles that seen in patients with primary depression. The isolation of patients meeting DSM-III criteria for conversion disorder yielded very similar results. Outcome and mortality data clearly separated conversion disorder from Briquet's syndrome patients; family history data suggested substantial diagnostic heterogeneity. Until the validity of this diagnosis is established, the label "conversion disorder" is recommended as a descriptive alternative to the label "undiagnosed."