Background and Aim: Rapid and sensitive detection of atrial fibrillation (AF) is of paramount importance for initiation of adequate preventive therapy after stroke. Stroke Unit care includes continuous electrocardiogram monitoring (CEM) but the optimal exploitation of the recorded ECG traces is controversial. In this retrospective single-center study, we investigated whether an automated analysis of continuous electrocardiogram monitoring (ACEM), based on a software algorithm, accelerates the detection of AF in patients admitted to our Stroke Unit compared to the routine CEM. Methods: Patients with acute ischemic stroke or transient ischemic attack were consecutively enrolled. After a 12-channel ECG on admission, all patients received CEM. Additionally, in the second phase of the study the CEM traces of the patients underwent ACEM analysis using a software algorithm for AF detection. Patients with history of AF or with AF on the admission ECG were excluded. Results: The CEM (n = 208) and ACEM cohorts (n= 114) did not differ significantly regarding risk factors, duration of monitoring and length of admission. We found a higher rate of newly-detected AF in the ACEM cohort compared to the CEM cohort (15.8% versus 10.1%, P < .001). Median time to first detection of AF was shorter in the ACEM compared to the CEM cohort [10 hours (IQR 0–23) versus 46.50 hours (IQR 0–108.25), P < .001]. Conclusions: ACEM accelerates the detection of AF in patients with stroke compared with the routine CEM. Further evidences are required to confirm the increased rate of AF detected using ACEM.