Abstract Background New-onset postoperative atrial fibrillation (POAF) complicates 1 in 3 cardiac surgeries and is associated with morbidity, mortality and clinical AF in long-term follow-up. Clinical risk scores have modest performance for predicting POAF. Polygenic risk scores are derived from the summation of up to millions of genetic variants and have shown good predictive ability for incident AF in the community. The ability of polygenic risk scores to predict POAF and subsequent recurrence of clinical AF in cardiac surgery patients is unclear. Methods We performed a prospective cohort study of patients from 4 regions (Canada, Hong Kong, Malaysia, United Kingdom) without a pre-operative history of atrial fibrillation (AF) who underwent cardiac surgery and were followed for 1 year. From pre-operative blood samples, we extracted DNA and calculated each participant’s polygenic risk score for AF using a penalized regression method (lassosum) to combine the effects of 5,000,621 genetic variants, weighted by their association with AF status from a previous genome-wide association study by Miyazawa (Nature Genetics, 2023). We estimated the association of this polygenic risk score for AF with the incidence of new-onset POAF using analyses adjusted for genetic ancestry. We assessed the ability of the polygenic risk score to predict POAF when added to common clinical risk scores. As a secondary objective, among patients who developed POAF, we estimated the association of the polygenic risk score with AF recurrence in follow-up beyond 30 post-operative days. Results Among 3031 patients (63.5% isolated coronary artery bypass grafting), 1282 patients (42.3%) developed new-onset POAF. The polygenic risk score for AF was strongly associated with the risk for POAF (odds ratio 1.3 per standard deviation increase in polygenic risk score [95% CI 1.2-1.4]). The 10% of participants with highest polygenic risk had a risk of POAF of 50.5% as compared to 41.4% for the bottom 90% (odds ratio 1.4 [95% CI 1.1-1.8]). When the polygenic risk score was added to the clinical risk scores, it improved the model fit for all scores, significantly improved the C-statistic for the CHA2DS2-VASc, POAF and HATCH Scores and improved measures of risk classification for all scores (Table). Follow-up data on AF status beyond 30 days were available for 902 patients; 71 patients (7.9%) had AF recurrence detected beyond 30 days post-operatively. The polygenic risk score was not significantly associated with a higher risk for AF recurrence (odds ratio, 1.1 per standard deviation increase in polygenic risk score [95% CI, 0.9-1.5]). Conclusions A higher polygenic risk score for AF is associated with the development of new-onset POAF following cardiac surgery and improves risk classification compared with clinical risk scores alone. However, this study failed to demonstrate an association of the polygenic risk score with AF recurrence in patients who develop POAF.