Abstract Aim To assess the risk of stroke and thromboembolism in patients with atrial fibrillation (AF) based on risk factor combinations of the CHA2DS2-VASc score. Methods and results Using nationwide Danish registries, patients with AF were included from 1997 to 2015 in this retrospective observational study. A multiple logistic regression, including interactions of history of stroke with age at AF, calendar year of AF, and the CHA2DS2-VASc score risk factors (congestive heart failure, hypertension, diabetes, vascular disease, and female sex) were used to predict the personalized risks of stroke within 1 year. A total of 147 842 patients with AF were included in the study cohort (median age 76 years, range 20–100 years, 51% females). Within the first year, 6% of the cohort were diagnosed with stroke. The predicted personalized 1-year absolute risk of stroke varied widely within each CHA2DS2-VASc score. To estimate the personalized risk of stroke an online calculator was created, the Calculator of Absolute Stroke Risk (CARS), which allows calculation of all the possible combinations of the CHA2DS2-VASc score (https://hjerteforeningen.shinyapps.io/riskvisrr/). Conclusion Calculation of the individual risk using a risk factor-based approach as opposed to using average risk for a particular CHA2DS2-VASc score can improve risk estimates. Furthermore, CARS can assist in the communication of the stroke risk for a more evidence-based shared decision-making of whether to initiate oral anticoagulation therapy.