Purpose: To investigate the hypothesis that sleep bruxism (SB) events could be predicted by an increase in heart rate. Materials and Methods: Fourteen sleep bruxers were recruited. Each participant recorded his or her own electromyography (EMG) and electrocardiography (ECG) at home for 2 consecutive nights using a portable telemetry system. Ten heartbeats before (B10 to B1) and three heartbeats after (A1 to A3) the onset of SB events were analyzed, and the threshold for the prediction of an SB event was determined. The validity of the threshold was tested by EMG and ECG recorded in the same manner for an additional night. The prediction accuracy of SB events was evaluated for sensitivity and specificity. Results: A gradual increase in heart rate was observed before an SB event, and B1, A1, A2, and A3 were significantly higher than B10 (P .01). The threshold value was set at 110% when the mean of all heart rates of the second night of recording was set at 100%. A total of 324 SB events were observed and 299 were preceded by increased heart rate that exceeded the threshold (sensitivity, 92.3%). The total number of increased heart rate events was 1,239, and the total number of threshold applications was estimated to be 120,000. The specificity was 99.2%. Conclusion: Over 90% of SB events could be predicted by an increasing heart rate of 110%. Since the sensitivity and specificity were extremely high, the hypothesis that SB events could be predicted by increased heart rate was positively verified.