Disuse atrophy and the accompanying loss of muscle strength with ageing is one of the causal factors of dysphagia (difficulty in swallowing). Evaluating the number of spontaneous swallowing during daily life is essential to detect the high-risk patient of disuse atrophy. We therefore developed an algorithm which automatically detects the number of spontaneous swallowing from single channel Electromyogram (EMG). Potential differences between electrodes attached on the thyroid cartilage and the lower tip of the chin were subjected to the analysis. We used individually-averaged EMG waveform as a template to detect swallowing-related EMG pattern since muscle activity during swallowing varies among subjects. We measured 3-5 times of spontaneous swallowing activity at the beginning of the measurement, which were averaged to generate individual template. We took a convolution between the raw EMG data and the individual template to find the waveform pattern which represents swallowing activity. Our preliminary experiment also suggested that the most frequent artifact was neck motion-related high-amplitude activity. These large artifacts were therefore automatically determined and rejected in the proposed algorithm. The feasibility of the proposed algorithm was tested using EMGs from two young-adult subjects without any problem with swallowing. Subjects sat in a chair for 30 min and watched a movie quietly. Experimenter counted the number of spontaneous swallowing by monitoring the movement of the thyroid cartilage of the subject via video monitor. The proposed algorithm successfully detected all swallowing events and there were no erroneous decision. The number of swallowing events detected by the proposed program was the same as those that were manually counted from visual inspection of the video. Our method would be useful in clinical approach to evaluate the number of swallowing in patients with high risk of dysphagia.