Detection of atrial fibrillation in patients presenting with ischemic stroke or transient ischemic attack (TIA) is important for the prevention of future events. We sought to develop a scoring system that would identify those patients most likely to have atrial fibrillation.Records from an inpatient stroke and TIA database and echocardiographic data were reviewed. Consecutive acute stroke and TIA patients over the age of 18 who were admitted during a two-year period were studied. Univariate and multivariable analyses were performed to identify variables associated with atrial fibrillation. Logistic regression analyses were used to develop a scoring system for atrial fibrillation.953 patient charts were reviewed; 145 patients (15%) had atrial fibrillation. In univariate and multivariate analyses, variables that were significantly associated with atrial fibrillation included left atrial diameter, age, and diagnosis of stroke. A history of smoking showed an inverse association. A 6-point scoring system based on these variables (with the acronym of LADS) was developed. A score of 4 or greater was associated with a sensitivity of 85.5% and a specificity of 53.1%. Approximately 47% of stroke and TIA patients would be excluded from further investigation using this score.We describe a system of scoring that identifies acute stroke and TIA patients with a greater chance of having atrial fibrillation. An inverse relationship with smoking was found. Further prospective studies are required to determine the clinical utility and cost-effectiveness of this scoring system in clinical practice and to investigate the inverse relationship between smoking and atrial fibrillation in this population.