Objective To explore the predictive value of heart rate variability(HRV)for long-term prognosis in elderly patients with stable coronary heart disease(CHD)based on dynamic electro-cardiogram(ECG).Methods A retrospective trial was conducted on 402 elderly patients with sta-ble CHD admitted to our hospital from January 2015 to December 2018,and all of them were fol-lowed up for 5 years.According to occurrence of major adverse cardiovascular events(MACE)or not,they were divided into a MACE group(n=102)and a control group(n=300).The main clin-ical characteristics and differences in HRV were compared between the two groups.Multivariate logistic regression analysis and ROC curve analysis were used.Results Advanced age,larger ratio of diabetes,more severe coronary artery stenosis,and higher low-frequency power and ratio of low-frequency power/high-frequency power,but lower left ventricular ejection fraction(LVEF)and high-frequency power were observed in the MACE group than the control group(P＜0.01).There was also statistical difference in use of hypoglycemic drugs between the two groups during the follow-up period(P＜0.01).Multivariate logistic regression analysis showed that diabetes,age ≥80 years,LVEF＜50％,coronary artery stenosis ≥70％and ratio of low-frequency power/high-frequency power＞1.16 were independent influencing factors for MACE in elderly patients with stable CHD within 5 years(P＜0.05,P＜0.01).ROC curve analysis indicated that the AUC value for low-frequency power and ratio of low-frequency power/high-frequency power in predic-ting MACE occurrence within 5 years was 0.801(95％CI:0.749-0.854,P=0.000)and 0.798(95％CI:0.752-0.844,P=0.000),respectively,and the value of high-frequency power in predic-ting the absence of MACE was 0.629(95％CI:0.566-0.692,P=0.000).Conclusion HRV is an independent influencing factor for MACE occurrence within 5 years in elderly patients with stable CHD,and it shows certain predictive value for MACE occurrence within 5 years in these patients.