The aim of this study was to identify differences in the medical management and clinical outcome in a group of elderly patients admitted to a designated geriatric assessment unit (GAU) or to two general medical units (GMUs). A prospective randomised controlled trial was undertaken in 267 patients aged 70 years and over (mean age = 78.3 years). Following discharge from hospital, patients were followed up at three monthly intervals for a total of 12 months. At the time of discharge, no significant differences were found in inpatient management, length of stay, mortality rates, discharge rates to institutional care or utilisation of community services in patients admitted to the GAU and the GMUs. Similarly, no significant differences were found at three, six, nine, and 12 month follow up in case fatality, activities of daily living indices, mental health status, rates of institutional referral and the level of community service support in patients admitted to the GAU and the GMUs studied. These findings do not show any advantage for the unselected 70+ acutely ill elderly patient who is admitted to a designated geriatric assessment unit rather than to a general medical unit. Therefore, an admission policy to GAU, based solely on age 70+ is medically inappropriate and cost-inefficient. Evidence from other sources suggests that an age cohort of acutely admitted patients beyond 80 years may well have returned more optimistic findings for the GAU. In future, GAUs will require a more selective admission policy to maximise the benefits of their rehabilitative and interdisciplinary approach.