We examined how the time and frequency domain measures of heart rate and blood pressure variability at supine rest reflect the sympathovagal balance of 23 female and male endurance athletes. Pharmacological blocking by atropine and propranolol was used as a standard for defining autonomic control of the heart. The Rosenblueth and Simeone model for neural control of heart rate was used to calculate the sympathovagal balance index (Abal). Atropinization significantly decreased all time and frequency domain measures of heart rate and blood pressure variability. beta-Blockade significantly decreased further the low- (< 0.07 Hz) and medium-frequency power (0.07-0.15 Hz) variability of R-R intervals (RRI) and SD of RRI. Abal was 0.629 +/- 0.019, indicating that parasympathetic activity predominated in the athletes. Basal heart rate (r = 0.519, P < 0.01), SD of RRI (r = -0.533, P < 0.01), root-mean-square of successive RRIs (RRI RMSSD) (r = -0.579, P < 0.05), RRI total (r = -0.557, P < 0.01) and RRI high-frequency (HF) power (r = -0.582, P < 0.01) correlated significantly with Abal and parasympathetic activity index. We concluded that the best non-invasive method of evaluating the sympathovagal balance of athletes at supine rest is to measure SD of RRI, RRI RMSSD, HF and total power of RRI variability. All heart rate variability measures were mainly parasympathetically modulated. The nature of blood pressure variability measures remained unclear and they could not be used to evaluate the sympathovagal balance among athletes.