The present study investigates the association of the predicted CHD-risk (PROCAM) with the individual endurance capacity and heart rate variability (HRV) in a population-based sample of sedentary elderly. After stratification, in 57 men (48.1+/-9.5 yrs.) with an overall PROCAM-risk <10% (28.7+/-10.9 points) and 22 men (54.5+/-7.7 yrs.) with a coronary 10-year risk > or =10% (50.8+/-5.6 points) cycle ergometries and short-term HRV analysis of time (RRMEAN, SDNN, RMSSD) and frequency domain parameters (LF, HF, TP, LF/HF) were conducted. Additionally the autonomic stress index (SI) was calculated. Nonparametric tests were used for statistical correlation analysis (Spearman rho) and group comparisons (Mann-Whitney). For endurance capacity [W/kg] (r=-0.469, p<0.001), SDNN (r=-0.302, p<0.05), RMSSD (r= -0.311, p<0.05), LF (r=-0.325, p<0.05), HF (r= -0.311, p<0.05) and TP (r= -0.307, p<0.05) negative monotone correlations with the coronary score-risk were determined. Significant positive correlations were calculated for SI (r=0.476, p<0.001). Except for RRMEAN and LF/HF significant group differences (p<0.05) were computed for SDNN (30.0+/-20.0 vs 20.0+/-10.0 ms), RMSSD (22.2+/-18.3 vs 18.0+/-8.7 ms), LF (90.9+/-241.5 vs 41.35+/-81.1 ms(2)), HF (43.0+/-105.1 vs 18.0+/-27.0 ms(2)) and TP (189.0+/-457.1 vs 100.0+/-157.6 ms(2)). Significant differences (p<0.01) were evaluated for exercise capacity (2.4+/-0.5 vs 1.8+/-0.3 W/kg) and SI (90+/-183 vs 322+/-291). The results underline the predictive value of HRV analysis in risk stratification and outline the interrelation of a decreased exercise capacity and autonomic function with a raised individual 10-year cardiac risk. As an independent parameter of the vegetative regulatory state the stress index may contribute to an increased practical relevance of short-time HRV analysis.