A study on the feasibility of obtaining usable electrocardiogram (ECG) signals from a wearable armband during 24-hour continuous monitoring is presented. The wearable armband records 3-channel ECG and, unlike the conventional Holter monitors, it is convenient for long-term daily life monitoring because it uses no obstructive leads and it is based on dry (no gels) electrodes, which do not cause skin irritation. An optimal channel selector is presented, based on a linear classifier using features that are related to the ECG signal quality. In addition, this linear classifier is also used for artifact detection. The developed optimal channel selector and artifact detector are applied to 24-hour armband ECG recordings from 5 subjects. For reference comparison, the subjects also wore a Holter device. The armband obtained usable data during 51.07±13.54% (inter-subject mean ± standard deviation) of the non-bed recording time, and the mean heart rate was accurately (relative error with respect to the Holter less than 10%) estimated from the armband selected ECG channel from 94.39±3.41% of the usable data. During the bed recording time, the percentage of usable data was 93.54±2.92%, and mean heart rate was estimated accurately from 97.01±1.80% of those data. These results suggest that the armband device is potentially feasible for a long-term daily life heart rate monitoring based on the presented channel selector and artifact detector, especially during the bed time.