Background It is unknown whether atrial fibrillation (AF) detected after acute ischemic stroke is caused by neurogenic or cardiogenic mechanisms. Based on the potential damage to the autonomic nervous system, neurogenic mechanisms could be implicated in the pathophysiology of newly diagnosed AF. To test this hypothesis, we developed a mechanistic approach by comparing a prespecified set of indicators in acute ischemic stroke patients with newly diagnosed AF, known AF, and sinus rhythm. Methods We prospectively assessed every acute ischemic stroke patient undergoing continuous electrocardiographic monitoring from 2008 through 2011. We compared newly diagnosed AF, known AF, and sinus rhythm patients by using 20 indicators grouped in 4 domains: vascular risk factors, underlying cardiac disease, burden of neurological injury, and in-hospital outcome. Results We studied 275 acute ischemic stroke patients, 23 with newly diagnosed AF, 64 with known AF, and 188 with sinus rhythm. Patients with newly diagnosed AF had a lower proportion of left atrial enlargement (60.9% versus 91.2%, P = .001), a smaller left atrial area (22.0 versus 26.0 cm2, P = .021), and a higher frequency of insular involvement (30.4% versus 9.5%, P = .017) than participants with known AF. Compared with patients in sinus rhythm, those with newly diagnosed AF had a higher proportion of brain infarcts of 15 mm or more (60.9% versus 37.2%, P = .029) and a higher frequency of insular involvement (30.4% versus 7.3%, P < .001). Conclusions The low frequency of underlying cardiac disease and the strikingly high proportion of concurrent strategic insular infarctions in patients with newly diagnosed AF provide additional evidence supporting the role of neurogenic mechanisms in a subset of AF detected after acute ischemic stroke.