This review article discusses the following, as yet unanswered, questions and research priorities to optimise patient management and stroke prevention in atrial fibrillation with the new direct oral anticoagulants (NOACs): 1. In patients prescribed a NOAC, can the anticoagulant effects or plasma concentrations of the NOACs be measured rapidly and reliably and, if so, can "cut-off points" between which anticoagulation is therapeutic (i.e. the "therapeutic range") be defined? 2. In patients who are taking a NOAC and bleeding (e.g. intracerebral haemorrhage), can the anticoagulant effects of the direct NOACs be reversed rapidly and, if so, can NOAC-associated bleeding and complications be minimised and patient outcome improved? 3. In patients taking a NOAC who experience an acute ischaemic stroke, to what degree of anticoagulation or plasma concentration of NOAC, if any, can thrombolysis be administered safely and effectively? 4. In patients with a recent cardioembolic ischaemic stroke, what is the optimal time to start (or re-start) anticoagulation with a NOAC (or warfarin)? 5. In anticoagulated patients who experience an intracranial haemorrhage, can anticoagulation with a NOAC be re-started safely and effectively, and if so when? 6. Are the NOACs effective and safe in multimorbid geriatric people (who commonly have atrial fibrillation and are at high risk of stroke but also bleeding)? 7. Can dose-adjusted NOAC therapy augment the established safety and efficacy of fixeddose unmonitored NOAC therapy? 8. Is there a dose or dosing regimen for each NOAC that is as effective and safe as adjusted-dose warfarin for patients with atrial fibrillation who have mechanical prosthetic heart valves? 9. What is the long-term safety of the NOACs? © Schattauer 2014.