PURPOSE: It has been shown that the power spectral density (PSD) of heart rate variability (HRV) can be decomposed into a power-law function and a residual PSD (rPSD) with a more prominent high-frequency component than that in traditional PSD. This study investigated whether the residual HRV (rHRV) measures can better discriminate patients with acute myocardial infarction (AMI) from patients with patent coronary artery (PCA) than traditional HRV measures. MATERIALS AND METHODS: The rHRV and HRV measures of 48 patients with AMI and 69 patients with PCA were compared. RESULTS: The high-frequency power of rHRV spectrum was significantly enhanced while the low-frequency and very low-frequency powers of rHRV spectrum were significantly suppressed, as compared to their corresponding traditional HRV spectrum in both groups of patients. The normalized residual high-frequency power (nrHFP = residual high-frequency power/residual total power) was significantly greater than the corresponding normalized high-frequency power in both groups of patients. Between-groups comparison showed that the nrHFP in AMI patients was significantly smaller than that in PCA patients. Receiver operating characteristic curve analysis showed that the nrHFP or nrHFP + normalized residual very low-frequency power (residual very low-frequency power/rTP) had better discrimination capability than the corresponding HRV measures for predicting AMI. CONCLUSIONS: Compared with traditional HRV measures, the rHRV measures can slightly better differentiate AMI patients from PCA patients, especially the nrHFP or nrHFP + normalized residual very low-frequency power.