Background: In most elderly people, the final-terminal-phase of life is characterized by permanent dependency and a complete inability to perform activities of daily living. Treatment targets usually switch from rehabilitation to palliation. However, the prevalence of the clinical judgment "last phase of life" and its association with in-hospital death is unknown in geriatric patients., Patients and Methods: We retrospectively analyzed GEMIDAS data from two geriatric units. Patients without cancer and an in-hospital stay of at least 1 week were included in our study. Prevalence of the terminal phase of life was clinically assessed according to the proposals made by M. Gillick. This clinical judgment was pronounced by the geriatric team after a stay in the hospital of at least 1 week. The clinical judgment took into account all available assessment parameters, as well as the impact of a geriatric treatment trial. In addition, the association between the clinical judgment and the risk of in-hospital mortality was analyzed., Results: Records from 2,433 (56%) patients in hospital A and from 1,912 (44%) patients in hospital B were analyzed. The frequency of a terminal phase of life was 30% and 9% (p<0.01), respectively. The frequency depended on the manner of admission to the hospital. In both hospitals, mortality was significantly higher in terminal patients (27% and 37%) than in other patients (0-8% and 0-6%). In both hospitals, the risk of in-hospital mortality was significantly associated with the clinical judgment (OR 3.1 and 2.7), heart failure (OR 2.2 and 2.1), and dementia (OR 2.0 and 1.8). Age, residency in a nursing home, and the Barthel Index on admission were all without relevant impact., Conclusion: The frequency of the clinical construct "terminal phase of life" varies in geriatric units between 9% and 30%. This clinical construct is significantly associated with increased in-hospital mortality. Therefore, this construct possesses external validity. Further studies are needed in order to assess the significance of such a clinical judgment, the associations with clinical burdens of symptoms, and the supply structure required to cover the needs of patients and their families.