Study objective: Syncope is a common disposition challenge for emergency physicians. Among the riskstratification instruments available, only the San Francisco Syncope Rule is rigorously developed. We evaluate its performance in Canadian emergency department (ED) syncope patients. Methods: This retrospective review included patients aged 16 years or older who fulfilled the definition of syncope (transient loss of consciousness with complete recovery) and presented to a tertiary care ED during an 18-month period. We excluded patients with ongoing altered mental status, alcohol/illicit drug use, seizure, and head and severe trauma. Patient characteristics, 5 predictors for the rule (history of congestive heart failure, hematocrit level 30%, abnormal ECG characteristics, shortness of breath, and triage systolic blood pressure 90 mm Hg), and outcomes (as per the original study) were extracted. Results: Of 915 visits screened, 505 were included. Forty-nine (9.7%) visits were associated with serious outcomes. The rule performed with a sensitivity of 90% (44/49 outcomes; 95% confidence interval [CI] 79% to 96%) and a specificity of 33% (95% CI 32% to 34%). Including monitor abnormalities in the ECG variable would improve sensitivity to 96% (47/49 outcomes; 95% CI 87% to 99%). Although physicians failed to predict 2 deaths, the rule would have predicted all 3 deaths that occurred after ED discharge. Implementing the rule in our setting would increase the admission rate from 12.3% to 69.5%. Conclusion: In this retrospective Canadian study, the San Francisco Syncope Rule performed with comparable sensitivity but significantly poorer specificity than previously reported. Implementing the rule would significantly increase admission rates. Further studies to either refine the San Francisco Syncope Rule or develop a new rule are needed. © 2009 by the American College of Emergency Physicians.