This study evaluated whether small shifts in cardiac autonomic balance with standing, as measured by heart rate variability (HRV), were prospectively associated with incident coronary heart disease (CHD) and mortality. Both Black and White men and women aged 45-64 years from the Atherosclerosis Risk in Communities Study (n = 9,267) were followed from 1987 to 1997 for myocardial infarction (n = 296), fatal CHD (n = 63), and non-CHD mortality (n = 533). HRV indices and mean R-R interval length (inverse of heart rate) were measured in the supine and standing positions for 2 minutes each; HRV shift was calculated as the difference between positions. After adjustment for demographic characteristics and medication use, HRV in each position was significantly inversely related to events in Cox proportional hazards models. With the exception of R-R interval length shift and myocardial infarction (hazard ratio = 1.42, 95% confidence interval: 1.02, 1.98 for the smallest vs. the largest quartile), there was no association between HRV shift and the other events. Despite clinical research suggesting that HRV shift with standing is a more sensitive measure of autonomic balance than is HRV in one position, simple measures such as heart rate change and supine and standing HRV were better predictors of events.