Frailty is common in patients with atrial fibrillation and may impact on antithrombotic and anti-arrhythmic treatment. To describe differences in clinical characteristics, prescription of antithrombotic and anti-arrhythmic medications and incidence of haemorrhage and stroke, between frail and non-frail older inpatients. Prospective observational study in patients aged ≥65 years with atrial fibrillation admitted to a teaching hospital in Sydney, Australia. Frailty was assessed using the Reported Edmonton Frail Scale, stroke risk with CHA2DS2-VASc score and bleeding risk with HAS-BLED score. Participants were followed after 6 months for haemorrhages and strokes. We recruited 302 patients (mean age 84.7 ± 7.1 years, 53.3% frail, 50% female, mean CHA2DS2-VASc 4.61 ± 1.44, mean HAS-BLED 2.97 ± 1.04). Frail participants were older and had more co-morbidities and higher risk of stroke but not haemorrhage. Upon discharge, 55.7% participants were prescribed with anticoagulants (49.3% frail, 62.6% non-frail, P = 0.02). Thirty-three per cent received antiplatelets only and 11.1% no antithrombotics, with no difference by frailty status. For anti-arrhythmics, 52.6% received rate-control drugs only, 11.8% rhythm-control drugs only and 13.5% both and 22.1% were not prescribed either, with no difference by frailty status. On univariate logistic regression, frailty decreased the likelihood of anticoagulant prescription (odds ratio (OR) 0.58, 95%CI 0.36-0.93), but this was not significant on multivariate analysis (OR 0.66, 95%CI 0.40-1.11). After 6 months, overall incidence of ischaemic stroke was 2.1%, and in patients taking anticoagulants, incidence of major/severe bleeding was 6.3%, with no significant difference between frailty groups. Frailty status had little impact on antithrombotic prescription and no impact on anti-arrhythmic prescription.