This study examines the efferent mechanisms of carotid sinus baroreflex influence on ventricular repolarization and refractory period compared with effects on atrioventricular (AV) nodal conduction. Pressure was controlled in both carotid sinuses by the Moisejeff technique in 16 chloralose-anesthetized dogs. Increases in carotid sinus pressure during pacing produced graded prolongation of AV nodal conduction, ventricular repolarization and refractory period with a threshold at a carotid sinus pressure of 120 mm Hg and a peak response at 200 mm Hg. Atropine, 0.4 mg/kg, attenuated the peak percent change in ventricular repolarization interval by only 12 +/- 14% (+/- standard error of the mean) despite a significantly greater attenuation (48 +/- 11%, p less than 0.05) in peak percent change in AV nodal conduction. However, stellate ganglionectomy attenuated the peak percent change in ventricular repolarization (42 +/- 19%), similar to effects on AV nodal conduction (59 +/- 21%, p greater than 0.25). Changes in mean arterial pressure, ventricular end-diastolic segment length or segment length shortening with systole (sonomicrometer technique) did not account for the electrophysiologic responses. Latency to peak effect on ventricular repolarization (43 +/- 7 seconds) was slower than that on AV nodal conduction (23 +/- 6 seconds, p less than 0.05). This difference in time course was not abolished by atropine. Thus, the carotid sinus baroreflex prolongs ventricular repolarization and refractoriness mainly by withdrawal of sympathetic influence; AV nodal conduction is prolonged by both vagal activation and sympathetic withdrawal. In addition, differences in time course between ventricular and AV nodal electrophysiologic responses are not explained by different efferent autonomic mechanisms.