BACKGROUND Dynamic cerebral autoregulation (CA) is known to be impaired in patients with acute ischemic stroke (AIS), but whether or not dynamic CA can predict long-term outcomes is unclear. METHODS This prospective study included 103 patients with AIS between September 2017 and April 2019. We measured the middle cerebral artery blood flow velocity and blood pressure within 7 days of AIS onset using a transcranial Doppler and Finometer, respectively. We conducted transfer function analysis to calculate dynamic CA indices (phase and gain), with lower phase and higher gain parameters reflecting less efficient CA. We followed up all patients after 3 and 12 months. Patients with 12-month modified Rankin Scale scores of <2 and ≥2 were defined as having favorable and unfavorable outcomes, respectively. We then analyzed the predictors of unfavorable outcomes after 3 and 12 months using logistic regression. RESULTS The ipsilesional phase parameter was significantly lower in patients with unfavorable outcomes than in those with favorable outcomes. Multiple logistic regression analysis revealed that the ipsilesional phase parameter and the National Institutes of Health Stroke Scale score were nonmodifiable predictors of short-term and long-term outcomes. Moreover, in receiver operating characteristic analysis, the area under the curve of the ipsilesional phase parameter was 0.646 (95% confidence interval: 0.513-0.779, P =0.044). Notably, the optimal cut-off value was 20.33 degrees (sensitivity: 63%, specificity: 70%). CONCLUSIONS Dynamic CA is an independent predictor of outcomes at 3 and 12 months in patients with AIS.