Background: Antiarrhythmic drug (AAD) therapy for atrial fibrillation (AF) can be associated with both proarrhythmic and noncardiovascular toxicities. Practice guidelines recommend tailored AAD therapy for AF based on patient-specific characteristics, such as coronary artery disease and heart failure, to minimize adverse events. However, current prescription patterns for specific AADs and the degree to which these guidelines are followed in practice are unknown. Methods: Patients enrolled in the Get With The Guidelines–Atrial Fibrillation registry with a primary diagnosis of AF discharged on an AAD between January 2014 and November 2018 were included. We analyzed rates of prescription of each AAD in several subgroups including those without structural heart disease. We classified AAD use as guideline concordant or nonguideline concordant based on 6 criteria derived from the American Heart Association/American College of Cardiology/Heart Rhythm Society AF guidelines. Guideline concordance for amiodarone was not considered applicable, since its use is not specifically contraindicated in the guidelines for reasons such as structural heart disease or renal function. We analyzed guideline-concordant AAD use by specific patient and hospital characteristics, and regional and temporal trends. Results: Among 21 921 patients from 123 sites, the median age was 69 years, 46% female and 51% had paroxysmal AF. The most commonly prescribed AAD was amiodarone (38%). Sotalol (23.2%) and dofetilide (19.2%) were each more commonly prescribed than either flecainide (9.8%) or propafenone (4.8%). Overall guideline-concordant AAD prescription at discharge was 84%. Guideline-concordant AAD use by drug was as follows: dofetilide 93%, sotalol 66%, flecainide 68%, propafenone 48%, and dronedarone 80%. There was variability in rate of guideline-concordant AAD use by hospital and geographic region. Conclusions: Amiodarone remains the most commonly prescribed AAD for AF followed by sotalol and dofetilide. Rates of guideline-concordant AAD use were high, and there was significant variability by specific drugs, hospitals, and regions, highlighting opportunities for additional quality improvement.