Objectives:To study the outcomes of elderly patients in a high-dependency care unit and to evaluate the costs and benefits of a geriatric high-dependency unit (GHDU).Design:Prospective data collection and analysis.Setting:Geriatric high-dependency unit.Patients:One hundred fifty patients ≥70 yrs of age who had been admitted to the GHDU over a 10-month period were investigated during their treatment and rehabilitation.Measurement and Main Results:The patients' Acute Physiology and Chronic Health Evaluation (APACHE) II scores and Simplified Acute Physiology Scores (SAPS) were recorded. The APACHE II scores and SAPSs provided a close correlation with the patients' mortality (correlation coefficients were 0.97 and 0.92, respectively). The SAPS proved to have a better linear relationship with the elderly patients' mortality in comparison with APACHE II scores. Most of the elderly patients included in the study were suffering from multiple premorbid medical problems. Overall, the mortality rate up to 1 month after discharge from the hospital was 48%. For patients ranging in age from 70 to 84 yrs, the 1-month mortality was 39.6%; however, for patients ≥85 yrs of age, the 1-month mortality was 68.1%. The mortality ratio was 0.96 (for all patients), 0.88 (for those ages 70-84 yrs), and 1.05 (for those age 85 yrs and above). For patients with nil organ system failure, the mortality rate was 32%. For patients with one organ system failure, the mortality increased to 48%. For patients with two organ system failures, the mortality rate was 86%. Survival for patients with three or more organ system failures was unprecedented. Survivors and nonsurvivors were compared. Three poor-prognosis groups were identified: group 1, patients who had received preadmission cardiopulmonary resuscitation; group 2, patients with a recent history of malignant diseases; and group 3, patients who had been mechanically ventilated. All three groups had a significantly higher mortality than those without these factors (p< .05). Patients in the 85 yrs and above group had a significantly higher mortality rate than those in the 70- to 84-yr age group (p< .05). Patients with SAPS and APACHE II scores >20 and >30, respectively, had a poor prognosis. The geriatric outcome scoring system (GOSS) was used as the functional outcome test for the survivors. The GOSS has three components: activities of daily living, mobility status, and social condition. At 1 month after discharge, 66.7% of the survivors returned to their premorbid activities of daily living abilities, 79.5% maintained their mobility status, and 91.7% remained at the same social environment. No survivors deteriorated more than one grade in any of the three components measured by the GOSS. The severity-of-illness scores, percentage of mechanical ventilation utilization, mortality rate, length of GHDU stay, and total hospital stay were comparable with those of other intensive care units (ICUs). The cost of 1 GHDU bed-day was equivalent to 24% of 1 ICU bed-day.Conclusion:The prognostic information that we gathered from an unselected group of critically ill elderly patients is useful. The GHDU achieved treatment results similar to those achieved by an ICU and is therefore seen as an innovative way of treating critically ill elderly patients. High-dependency care for the elderly patient is worthwhile.