Dependence of QT interval duration on cardiac heart rate has been well established and is considered to be an intrinsic property of ventricular myocardium. Conclusive results of autonomic influences on such phenomena are lacking. To evaluate whether rate-dependent changes of QT interval are conditioned by the autonomic nervous system, 28 normal subjects with no heart disease and a normal QT interval were electrophysiologically assessed. The QT interval was calculated at 6 paced cycle lengths (600, 540, 500, 460, 430 and 400 ms) during the basal state, and after beta blockade (propranolol 0.2 mg/kg) and autonomic blockade (propranolol plus atropine 0.04 mg/kg). Because of atrioventricular nodal conduction limits, intrapatient cross-comparisons were performed in 10 subjects (aged 42 +/- 15 years). Single regression lines, evaluated in each subject, showing correlation between pacing cycle length and QT duration at each of the 3 states were analyzed. The mean slope observed after autonomic blockade (b = 0.10 +/- 0.04) was significantly lower than that seen during the basal state (b = 0.22 +/- 0.12, p less than 0.05) and after beta blockade (b = 0.23 +/- 0.08, p less than 0.05); nonsignificant differences were found between slopes during the basal state and after beta blockade. Results showed that vagal tone increased intrinsic dependence of QT at increasing cycle length, whereas sympathetic tone did not seem to interfere significantly. Since (in each subject) beta blockade was performed--or achieved--before atropine administration, the vagal influences are likely to be directly exerted on the ventricular electrophysiologic substrate.