The present study investigated the effects of long-duration exercise on heart rate variability (as a marker of cardiac vagal activity, VT). Heart rate variability (time series analysis) was measured in mongrel dogs (n=24) with healed myocardial infarctions during 1 hour of submaximal exercise (treadmill running at 6.4 kph/10% grade). Long-duration exercise provoked a significant (ANOVA, all P<0.01, mean±SD) increase in heart rate (1^st min 165.3±15.6 vs. last min 197.5±21.5 beats/min) and significant reductions in high frequency (0.24 to 1.04 Hz) power (VT, 1^st min 3.7±1.5 vs. last min 1.0±0.9 ln ms²), R-R interval range (1^st min 107.9±38.3 vs. last min 28.8±13.2 ms), and R-R interval SD (1^st min 24.3±7.7 vs. last min 6.3±1.7 ms). As endurance exercise training can increase cardiac vagal regulation, the studies were repeated after either a 10-week exercise training (n=9) or a 10-week sedentary period (n=7). After training, long-duration exercise elicited smaller increases in heart rate (pre-training, 1^st min 156.0±13.8 vs. last min, 189.6±21.9; post-training, 1^st min 149.8±14.6 vs. last min 172.7±8.8 beats/min) and that were accompanied by smaller reductions in heart rate variability (e.g., VT, pre-training, 1^st min 4.2±1.7 vs. last min 0.9±1.1; post-training, 1^st min 4.8±1.1 vs. last min 2.0±0.6 ln ms²). The response to long-duration exercise did not change in the sedentary animals. Thus, the heart rate increase that accompanies long-duration exercise results, at least in part, from reductions in cardiac vagal regulation. Furthermore, exercise training attenuated these exercise-induced reductions in heart rate variability, suggesting maintenance of a higher cardiac vagal activity during exercise in the trained state.