The objective of this study was to determine whether having a hospice unit within the hospital increases the proportion of terminally ill patients who use hospice services (including home, nursing home, or inpatient hospice) post-admission. Using medical record data abstracted for 232 randomly selected patients with terminal cancer admitted to six community hospitals in Connecticut, we found that patients admitted to a hospital with a hospice unit were more likely to use hospice services (i.e., home hospice, nursing home hospice, or inpatient hospice) post-admission than patients admitted to a hospital without a hospice unit (unadjusted OR 5.7, 95% CI 3.1, 10.6). This effect persisted after adjusting for patient age, gender, marital status, documented discussions of prognosis, prior hospice use, and type of cancer.