The purpose of this study was to investigate on the perception of a bodily symptom, 'accelerated heartbeat'. In this study, three experimental conditions were set up. First condition was the true-feedback (TF) condition, in which subject was presented true feedback signals corresponding to actual heartbeats. Second condition was the false-feedback (FF) condition, in which subject was presented the false feedback signals being irrelevant to actual heartbeats. Third condition was the no-feedback (NF) condition. The major results were as follows ; (1) In the TF-condition, abilities to perceive accelerated heartbeat were different inter- and intra-subjects. (2) In the FF-condition, almost subjects perceived accelerated heartbeat were inaccurate. (3) Frequencies of perception in the NF-condition were fewer than in the TF-condition. As the cue for perception of accelerated heartbeat, subjects used not only information relevant to heartbeats such as a pulse but also respiration, degree of concentration, and so on. These results suggested that abilities to perceive accelerated heartbeat were different inter- and intra-individuals, even when they could discriminate their heartbeats perfectly, and that, if an information unrelated to accelerated heartbeat was used as the cue, subject perceived their accelerated heartbeat inaccurately. Further, the cue for interoceptive perception of accelerated heartbeat was not restricted information directly relevant to heartbeats.