The purpose of this study was to examine the effects of behaviorally conditioned changes in autonomic activity on atrio-ventricular (AV) transmission in dog. To produce consistent activation of the cardiac nerves in the awake animal (n = 7), a classical appetitive conditioning paradigm was used. A conditioning trial consisted of a 30 s control period followed by one of two differing situations: (1) a 60-s conditional stimulus (CS +) tone wherein food (i.e. ‘UCS’ or unconditioned stimulus) was given during the last 30 s; or (2) at 30-s discriminative stimulus (CS −) tone which was never followed by food reward. Eight of each type trial were given daily until a stereotypic cardiovascular response was developed for the CS + but not the CS −. The hemodynamic conditional response (i.e. ‘CR’, the response to the CS +) consisted of a moderate tachycardia (+ 14.5%, P ≤ 0.05), a small pressor response (+6.7%, P ≤ 0.01), and a moderate increase in the first time derivative of left ventricular pressure (+14.9%, P ≤ 0.01) reflecting an increase in inotropic state. The unconditional response (i.e., ‘UCR’, the response to the food reward) consisted of a substantial increase in HR (25.7%, P ≤ 0.01) above CR values while left ventricular pressure (LVP) and d(LVP)/dt increased 5.0% and 10.0% ( P ≤ 0.01 for both) above their CR values. The effect of the conditioned changes in neural activity on the AV node was observed by pacing the atrium from 110 to 180 bpm during the first 15 s of each trial period (i.e. control, CS +, UCS). The discrepancy between the atrial pace rate and the transmitted ventricular rate is expressed as a ‘mean difference score’ and serves as an index of the fidelity of the AV transmission process: the smaller the difference, the closer a 1 : 1 ratio of atrial vs ventricular beats is approached. The relatively large mean difference score for the control periods (46.0 ± 9 bpm) indicates that the paced atrial impulse did not faithfully precede ventricular contraction during these periods. The mean difference significantly decreased (34.6%, P ≤ 0.05) during the CS +, and approached an almost 1 : 1 ratio (75.6% decrease from CS + values, P ≤ 0.01) during food delivery. β-Adrenergic blockade (propranolol, 1 mg/kg, i.v.) eliminated the changes in mean difference during the CS + but not during food delivery. There were no statistically significant physiological changes during CS −. The results indicate that behaviorally mediated increases in sympathetic tone during the CS + improve the fidelity of the AV transmission process, whereas decreased parasympathetic tone appears to contribute to the increased AV transmission during the food reward period. Finally, sino-atrial and atrio-ventricular nodal function appear to be regulated in concert by the autonomic nervous system.