A disruptive area for mobile health (mHealth) in cardiovascular disease relates to the ability of mobile and wearable devices to detect cardiac arrhythmia, with currently a particular focus on atrial fibrillation (AF)(1–3). It might also provide an opportunity for timely identification of acute cardiac disorders and patients at risk for malignant cardiac arrhythmias, and possibly allow an earlier diagnosis of cause of palpitations and syncope (4). In this context, the logical next step is to assess the ability of wearable sensors to accurately detect cardiac arrhythmias beyond AF, as well as other electrical abnormalities, which has not been evaluated so far. We therefore evaluated the extent to which a commercially available, popular wearable sensor, Kardia Band (KB)(AliveCor, Inc., Mountain View, California) was able to identify commonly encountered electrocardiographic abnormalities. We examined whether the KB could accurately and reliably differentiate normal rhythm from common, classic abnormalities, when compared with physician-interpreted 12-lead electrocardiograms (ECGs). This study was conducted in patients (age $18 years) hospitalized between April and September 2018 (elective and emergency), who agreed to participate, in the cardiology departments or in the cardiac intensive care units of 2 academic hospitals (European Georges Pompidou Hospital, Paris, and Poitiers University Hospital, Poitiers) in France, and funded by the French Society of Cardiology. Our study received the proper ethical oversight, with approval from our local institutional review board/ethics committee. All pairs of 12-lead ECG (10-s strip) and KB strip were successively obtained close to each other. Pairs considered being not concomitant, or if a patient’s rhythm appeared to have changed from the time the 12-lead ECG was obtained to the time the KB strip was taken, were excluded (29 patients, 2.2%). All recordings were independently and centrally analyzed by 2 blinded cardiologists, according to a systematic and standardized reading flowchart, after mixing them up randomly so that the reading cardiologist did not know the ECG corresponding to a particular KB strip. If the interpretation of one of the strips (12-lead ECG or KB strip) did not match between the 2 cardiologists, the strip in question was further analyzed by a third expert, namely a cardiac electrophysiologist. We determined, for each diagnosis, the sensitivity and specificity (and respective confidence intervals) of the KB-device strip, with the 12-lead ECG as reference. All tests were 2-tailed with an alpha of 5%. A total of 1,322 patients were included (mean age 68.7+ 16.2 years, 863 [65.3%] men). Rhythm disorders were recorded in 374 (28.3%) of ECGs, and 1,141 (86.3%) patients had at least 1 other ECG abnormality. For arrhythmias, sensitivity and specificity of the KB-device were 82% and 92% for AF diagnosed by physicians, 95% and 86% for AF diagnosed by the automated KB algorithm, 26% and 98% for other supraventricular tachycardia, and 60% and 100% for wide-QRS tachycardia, respectively. For conduction disorders, sensitivity and specificity of the KB-device were 70% and 100% for highdegree atrioventricular block, and 50% and 100% for sinoatrial block, respectively. For the diagnosis of ST-T changes, sensitivity and specificity of the KB-device were 88% and 82% for ST-segment depression, 33% and 97% for ST-segment elevation, and 56% and 95% for T-wave inversion, respectively.