Although recent-onset atrial fibrillation and flutter are common arrhythmias managed in the emergency department (ED), there is insufficient evidence to help physicians choose between 2 competing treatment strategies, rate control and rhythm control. We seek to evaluate variation in ED management practices for recent-onset atrial fibrillation and flutter patients at multiple Canadian sites and to determine whether hospital site was an independent predictor of attempted cardioversion. We conducted a cross-sectional survey by health records review on an observational cohort of all eligible adult recent-onset atrial fibrillation and flutter cases, with onset of symptoms less than 48 hours, treated at 8 academic hospital EDs during a 12-month period, and evaluated the variation in practice among sites for important management strategies. Among the 1,068 study patients, 88.3% had atrial fibrillation and 11.7% had atrial flutter. The proportion of cases managed with rhythm control was 59.4% (interhospital range 42% to 85%) and, among these, electrocardioversion was attempted first for 44.2% (range 7% to 69%). There was variation in most management strategies, including use of rate control drugs 54.9% (range 37% to 65%), choice of procainamide as rhythm control drug 62.1% (range 15% to 89%), referral to cardiology in the ED 30.7% (range 16% to 64%), use of heparin 13.7% (range 1% to 29%), and outpatient cardiology referral 43.0% (range 30% to 65%). Adverse events were relatively uncommon and transient for patients undergoing attempts at pharmacologic (13.0%) or electrocardioversion (12.1%). Overall, 83.3% of patients were discharged home from the ED (range 73% to 90%). After controlling for 12 covariates, multivariate logistic regression found that factors independently associated with attempted cardioversion were age (odds ratio [OR] 0.97; 95% confidence interval [CI] 0.95 to 0.98), history of electrocardioversion (OR 2.73; 95% CI 1.56 to 4.80), associated heart failure (OR 0.29; 95% CI 0.09 to 0.95), and hospital site (ORs ranged from 0.38 to 3.05). We demonstrated a high degree of variation in management approaches for recent-onset atrial fibrillation and flutter patients treated in academic hospital EDs. Individual hospital site, age, previous cardioversion, and associated heart failure were independent predictors for the use of rhythm control.