A national survey of hospices revealed that few patients with primary dementia are currently treated by these programs, unless they have other terminal illnesses. An inability to predict survival was the major reason offered to explain this phenomenon. The higher percentage of patients with secondary dementia in hospice suggests that dementia per se does not make hospices care unfeasible. Similarly, the high proportion of for-profit hospices that enrolled patients whose dementia was primary implies the fiscal feasibility of providing hospice care for these patients. Further study is needed concerning the determinants of survival time in end-stage dementia.