BACKGROUND AND PURPOSE: We compared the associations of self-reported atrial fibrillation (AF) and ECG-detected AF with incident stroke in the Risk of Stroke in the Reasons for Geographic and Racial Differences in Stroke (REGARDS) study. METHODS: In this analysis, 27 109 participants aged 45 years or older without previous stroke were included. Stroke cases were identified and adjudicated during an average of 4.4 years of follow-up. Cox proportional hazards analysis was used to calculate hazard ratios (HR) of self-reported AF, ECG-detected AF, and AF detected by either method with incident stroke. We also examined the predictive ability of the Framingham Stroke Risk Score (FSRS) when the component AF was defined by different methods. RESULTS: After adjustment for components of the FSRS, self-reported AF, ECG-detected AF, and AF by either method were predictive of incident stroke (HR, 1.41; 95% CI, 1.05-1.88; HR, 1.90; 95% CI, 1.10-3.27; HR, 1.53; 95% CI, 1.16-2.01, respectively). When self-report, ECG, or either method, separately, were considered as the method of AF ascertainment in the FSRS, the HR per 1% increase in the FSRS were identical across AF ascertainment methods (HR, 1.04; 95% CI, 1.03-1.04; HR, 1.04; 95% CI, 1.04-1.05; HR, 1.04; 95% CI, 1.03-1.04; respectively). CONCLUSIONS: Self-reported AF is a strong predictor of stroke that can be used interchangeably or in combination with ECG-detected AF in stroke risk prediction models.