Background: Anticoagulation therapy with anti-vitamin K (AVK; vitamin K antagonist) for the prevention of thromboembolism in patients with atrial fibrillation (AF) is based on information derived from numerous well constructed, randomized controlled trials. Despite this conclusive evidence of efficacy, several studies have shown that 'real world' use of AVK in patients with AF is suboptimal. Our hypothesis was that geriatric characteristics (functional impairment, cognitive disorders, malnutrition, risk for falls, depression) could be an explanation for the underprescription of AVK in older patients with AF., Objective: To analyse the barriers to the prescription of AVK therapy, with special attention on geriatric characteristics., Methods: This was a retrospective study of 768 consecutive geriatric patients admitted to an acute geriatric unit of an academic hospital between April 2006 and November 2008. Data from comprehensive geriatric assessments were collected from computerized medical charts., Results: Analysis of data from the 768 medical charts showed that 111 patients (14%) presented with AF. Among the 111 patients (72% women), 45% were living in an institution before admission. These patients presented a high prevalence of geriatric syndromes: cognitive disorders 59%, malnutrition risk 59%, incontinence 35%, depression 37%, and falls 61%. Ninety percent of the patients had an Identification of Seniors At Risk (ISAR) questionnaire score ≥2, which indicates an increased risk of frailty and functional decline during hospitalization. The prevalence of conditions measured by the CHADS(2) (congestive heart failure, hypertension, age >75 years, diabetes mellitus and previous stroke or transient ischaemic attack [TIA]) clinical prediction scale was as follows: heart failure 19%, hypertension 79%, age >75 years 95%, diabetes 15% and stroke 24%. The mean ± SD number of daily classes of drugs received at admission was 7.1 ± 3.3 (median 7, range 0-20). Forty-nine percent of patients had not received any AVK treatment before admission. The level of functional dependence for basic and instrumental activities of daily living did not differ between patients receiving AVK before admission and those not receiving AVK. Similarly, the proportion of geriatric problems (cognitive, malnutrition, depression and falls) did not differ between these two groups. To determine whether the decision to administer AVK therapy before admission was influenced by the risk of an embolic stroke, determined by the presence of CHADS(2) conditions, we compared the proportions of patients who fulfilled those conditions in each group: again, no difference was found., Conclusions: Almost half of the patients presenting with AF did not receive any AVK therapy before admission. In this population, in which most patients had multiple impairments, no single impairment or geriatric characteristic was identified as a barrier to AVK use. It is possible that combinations of impairments and geriatric characteristics were barriers to the prescription of AVK therapy across the whole of this population. More research is needed to identify and clarify the relative importance of patient-, physician- and healthcare system-related barriers to the prescription of AVK therapy in older patients with AF.