Heart rate variability has been demonstrated both experimentally and clinically to be of prognostic importance in determining mortality after myocardial infarction. However, no paired studies have been reported to examine heart rate variability before and after myocardial infarction. The hypothesis was tested that low values of heart rate variability provided risk assessment both before and after myocardial infarction with use of an established canine model of sudden cardiac death. Risk for sudden death was assessed 1 month after myocardial infarction by a protocol in which exercise and myocardial ischemia were combined; dogs that developed ventricular fibrillation were classified at high risk for sudden death (susceptible) and the survivors were considered low risk (resistant). In resistant dogs, myocardial infarction did not affect any measure of heart rate variability: 1) mean RR interval, 2) standard deviation of the mean RR interval, and 3) the coefficient of variance (standard deviation/RR interval). By contrast, after myocardial infarction, susceptible dogs showed significant decrease in all measures of heart rate variability. Before myocardial infarction, no differences were seen between susceptible and resistant dogs. However, 30 days after infarction, epidemiologic analysis of the coefficient of variance showed high sensitivity and specificity (88% and 80%, respectively), predicting susceptibility. Therefore, results of analysis of 30 min of beat to beat heart period at rest 30 days after myocardial infarction are highly predictive for increased risk of sudden death.