Alligators and other crocodilians have a cog-wheel valve located within the subpulmonary conus, and active closure of this valve during each heart beat can markedly and phasically increase resistance in the pulmonary outflow tract. If this increased resistance causes right ventricular pressure to rise above that in the systemic circuit, right ventricular blood can flow into the left aorta and systemic circulation, an event known as pulmonary-to-systemic shunting. To understand better how this valve is controlled, anaesthetized American alligators (Alligator mississippiensis) were used to examine the relationships between depolarization of the right ventricle, depolarization/contraction of the cog-wheel valve muscle and the resultant right ventricular, pulmonary artery and systemic pressures. Depolarization swept across the right ventricle from the apex towards the base (near where the cog-wheel valve muscle is located) at a velocity of 91+/-23 cm s(-1) (mean +/- S.E.M., N=3). The cog-wheel valve electrocardiogram (ECG) (and thus contraction of the valve) trailed the right ventricular ECG by 248+/-28 ms (N=3), which was equivalent to 6-35 % of a cardiac cycle. This long interval between right ventricular and valve depolarization suggests a nodal delay at the junction between the base of the right ventricle and the cog-wheel valve. The delay before valve closure determined when the abrupt secondary rise in right ventricular pressure occurred during systole and is likely to strongly influence the amount of blood entering the pulmonary artery and thus to directly control the degree of shunting. Left vagal stimulation (10-50 Hz) reduced the conduction delay between the right ventricle and cog-wheel valve by approximately 20 % and reduced the integrated cog-wheel ECG by 10-20 %. Direct application of acetylcholine (1-2 mg) also reduced the integrated cog-wheel ECG by 10-100 %; however, its effect on the conduction delay was highly variable (-40 to +60 %). When the cog-wheel valve muscle was killed by the application of ethanol, the cog-wheel ECG was absent, right ventricular and pulmonary pressures remained low and tracked one another, the secondary rise in right ventricular pressure was abolished and shunting did not occur. This study provides additional, direct evidence that phasic contraction of the cog-wheel valve muscle controls shunting, that nervous and cholinergic stimulation can alter the delay and strength of valve depolarization and that this can affect the propensity to shunt.