BACKGROUND Heart rate (HR) variability is a marker of tonic cardiac autonomic activity and contributes in assessing risk for sudden death after myocardial infarction. Recent clinical observations have indicated that attenuation of HR variability, which occurs after myocardial infarction, may be transient. This study addresses the issue of whether autonomic control of heart rate recovers at different rates after myocardial infarction in subjects at high and low risk for ventricular fibrillation (VF). RESULTS Thirty dogs, 22 with myocardial infarction and 8 sham-prepared animals, completed the study. Changes and recovery in cardiac autonomic activity after myocardial infarction were examined by measuring HR variability before and at defined intervals during the first 30 days after infarction. Each HR variability measurement was made before and after beta-blockade in dogs at high (n = 10) and low (n = 12) risk for VF. Arrhythmia risk was determined on the basis of development of VF during exercise and transient myocardial ischemia 30 days after infarction. No sham-prepared animals developed VF. Preinfarction measurements of HR variability were not different between the groups before beta-blockade, but HR variability increased much more in response to beta-blockade in animals destined to be resistant compared with susceptible animals (289 +/- 26 to 369 +/- 35 msec, delta 27.7%, versus 270 +/- 36 to 283 +/- 34 milliseconds, delta 4.8%, respectively, P < .01). Immediately after infarction, HR variability was significantly attenuated in all dogs, but in the resistant dogs it recovered to pre-myocardial infarction levels within 10 days. After the infarction, beta-blockade did not increase HR variability in either group of animals. Postoperative increases in HR variability from beta-blockade were preserved in the sham group. Susceptible animals were characterized by a persistent attenuation of HR variability throughout the 30 days. CONCLUSIONS The depression in HR variability produced by myocardial infarction has a clearly different temporal recovery pattern between low- and high-risk animals. After myocardial infarction, beta-adrenergic blockade does not alter HR variability, thus preserving its predictive value. Before myocardial infarction, however, beta-blockade increases HR variability only in the animals destined to be at low risk for lethal arrhythmias after the infarction. The recovery pattern of HR variability after myocardial infarction may contribute to the early recognition of individuals at high risk for sudden death.