Variability in swallowing electromyography (EMG) has been reported to occur due to individual differences, even in studies with appropriate control of factors affecting EMG such as food type, properties, and posture. The increased variability broadens the confidence intervals of measures obtained from EMG, such as time and amplitude, and makes it difficult to detect differences in muscle activity. This makes it impossible to detect slight changes in swallowing function at an early stage, which is required for clinical examination. In this study, we focused on the tipper type and dipper type of oral swallows, which differ in the timing of muscle activity. The frequency of occurrence of both types varies between participants, and both types could occur in the same participant. In this study, we measured swallowing EMG wherein participants were instructed to swallow with a tipper and a dipper. The results showed that there was a significant difference in the between-participant duration between these two types of swallows. Within-participant variability was significantly lower in the tipper-dipper instruction condition compared to the baseline condition, which assumed a mixture of tipper-dipper swallowing. These results demonstrate that instructions on swallowing method are effective in improving the reproducibility of EMG.