Decreased heart rate variability (HRV) predicts adverse outcomes. HRV can be elevated by episodes of significant non-respiratory sinus arrhythmia (i.e., a highly erratic sinus rhythm with normal p-waves, ESR). This elevated HRV could confound risk stratification by increasing HRV in high-risk patients. HRV was determined from tapes recorded at baseline in the Cardiovascular Health Study, a population study of older adults. Twenty-four hour time, frequency and non-linear domain HRV was compared between ESR+ and ESR- subjects, with (CVD+) and without (CVD-) cardiovascular disease. ESR+ was associated with higher HRV in the time and frequency domains and with decreased short-term fractal scaling exponent and increased ratios of the dimensions of the Poincaré plot fitted ellipse. (ESR+ and CVD+) subjects with had the highest HRV for virtually all indices, while (ESR- and CVD+) had the lowest HRV. Since decreased HRV is associated with adverse outcomes, ESR is likely to dilute the predictive power of HRV.