Introduction: Atrial fibrillation (AF) is the most common arrhythmia treated in the emergency department (ED) and is associated with an increased risk of ischemic stroke. Studies have shown that only oral anticoagulant (OAC) therapy reduces risk of AF related stroke. Our objective was to measure the prescribing practices for OACs for new onset AF at a tertiary ED and two surrounding community EDs, and identify rates of adverse effects within 90 days. The findings of this study will provide quality assurance information for the management of patients with new onset AF. This information has the potential to promote adherence to prescribing guidelines for AF in the ED and the reduction of common adverse events such as ischemic stroke. Methods: We conducted a retrospective chart review of 385 patients with new onset AF who presented to the ED between November 2014 to Mach 2018. We defined new onset as symptoms <48 hours and had AF confirmed with electrocardiogram. We recorded the selected therapy choice of cardioversion and/or rate control, gender, age, and assessed CHADS-65 score. We recorded who was prescribed OAC and those who were referred to cardiology, family medicine, or did not have a documented follow up plan. Patients with a previous history of AF or current anticoagulant therapy were excluded. We recorded if any patients returned to the ED within 90 days with ischemic stroke, AF recurrence, myocardial infarction, other embolic disease or death. Results: 86 of 294 (29.5%) of patients who qualified under CHADS-65 received OACs appropriately. 64 of 66 (97.0%) of patients who did not qualify under CHADS-65 did not receive OACs appropriately. 5 patients overall returned within 90 days with ischemic stroke, 4 of those were not prescribed OACs, however this was not statistically significant (P = 0.999). Conclusion: This data suggests that physicians in the study are under-prescribing OACs relative to published guidelines. A larger study is necessary to elucidate the effect of ED OAC prescribing patterns on long-term patient outcome.