This study was designed to assess autonomic effects on the QT interval during recovery from exercise. Exercise is associated with an acute increased risk of sudden cardiac death. Evidence of impaired parasympathetic activity, such as low heart rate variability and heart rate recovery, and an increased QT interval are also associated with increased mortality. However, there is no clear pathophysiological link among these findings. Bicycle exercise testing was performed serially in 33 healthy volunteers (19 men; ages, 54 +/- 7 yr) under four conditions: 1) baseline, 2) beta-adrenergic blockade-intravenous propranolol (0.2 mg/kg) administered during exercise, 3) parasympathetic blockade-intravenous atropine (0.04 mg/kg) administered during exercise, and 4) double blockade with propranolol and atropine. ECGs were obtained every minute in recovery for 10 min and then at the 15th and 20th min, from which the QT and RR intervals were measured. Linear regression analyses were used to assess the individual QT-RR relationships for each subject for each condition. Relative to baseline, the QT-RR relationship with parasympathetic blockade was shifted to the left and had a steeper slope. In contrast, the QT-RR relationship with beta-adrenergic blockade was shifted to the right and had a less steep slope. The baseline and double-blockade QT-RR relationships were in the middle and essentially superimposable. There was a negative relationship between QT-RR slope and heart rate or RR interval recoveries, but it was significant only for the 1- and 2-min RR interval recoveries with low R(2) values of 0.124 and 0.114. The main parasympathetic effect in the postexercise recovery period is to counteract the sympathetically mediated QT prolongation. These data support the concept that parasympathetic tone may provide a natural antiarrhythmic effect during this time.