The role of antithrombotic therapy in the prevention of ischemic stroke after non-cardiac surgery is unclear. In this study, we tested the hypothesis that the association of new-onset postoperative atrial fibrillation (POAF) on ischemic stroke can be mitigated by postoperative oral anticoagulation therapy. Of 251,837 adult patients (155,111 female (61.6%) and 96,726 male (38.4%)) who underwent non-cardiac surgical procedures at two sites, POAF was detected in 4,538 (1.8%) patients. The occurrence of POAF was associated with increased 1-year ischemic stroke risk (3.6% versus 2.3%; adjusted risk ratio (RRadj) = 1.60 (95% confidence interval (CI): 1.37-1.87), P < 0.001). In patients with POAF, the risk of developing stroke attributable to POAF was 1.81 (95% CI: 1.44-2.28; P < 0.001) without oral anticoagulation, whereas, in patients treated with anticoagulation, no significant association was observed between POAF and stroke (RRadj = 1.04 (95% CI: 0.71-1.51), P = 0.847, P for interaction = 0.013). Furthermore, we derived and validated a computational model for the prediction of POAF after non-cardiac surgery based on demographics, comorbidities and procedural risk. These findings suggest that POAF is predictable and associated with an increased risk of postoperative ischemic stroke in patients who do not receive postoperative anticoagulation.