Increased sympathetic tone with exercise enhances ventricular conduction and would be predicted to shorten QRS duration. Previous studies, however, have not consistently documented such changes. Using a digital electrocardiograph (ECG) cart sampling at 500 Hz, a bipolar precordial lead (V5-V2) was recorded (supine, at end expiration) in 25 healthy men (mean age 29 yr, range 19-37) at rest and immediately after submaximal treadmill exercise. QRS duration was measured on complexes recorded at high gain and expanded time scale. A significant (P less than 0.0005) decrease [4.9 +/- 2.3 (SD) ms] in QRS duration was noted, and decreased QRS duration was observed in all 25 subjects (range 1-10 ms). Furthermore there was significant (P less than 0.01) shortening of the intervals between QRS onset and Q-wave nadir (1.2 +/- 2.0 ms) and between QRS onset and R-wave peak (2.4 +/- 2.5 ms), as well as of the Q-R interval (1.3 +/- 2.7 ms). However, there were no significant differences in percent shortening of early vs. later phases of the QRS. To exclude digital sampling errors the same protocol was also applied in 11 healthy men by using an analog ECG cart. QRS duration immediately postexercise shortened by 7.1 +/- 4.3 ms (range 2-11.5 ms), which was not significantly different from the results obtained with the digital ECG. We conclude that decreased QRS duration is a physiological response to moderate treadmill exercise in healthy men, reflecting enhancement of conduction in early, middle, and later phases of ventricular activation.