Ninety-three rapid atrial pacing studies were performed in 38 children to compare preoperative and early postoperative function of the sinus and atrioventricular (AV) nodes. The interval between the preoperative and postoperative studies was under 6 months in the majority of patients. Postoperative studies were performed within 48 hours of operation and between 4 and 8 days after operation. Sinus nodal function as measured by sinus nodal recovery time (SNRT) was an unreliable index in determining depression since the number who improved postoperatively (10/55) was nearly equal to the number that worsened (12/55). The majority who had abnormal function postoperatively demonstrated a junctional rather than sinus recovery focus. This finding appears a more definitive and more reproducible indicator of sinus node depression in the postoperative patient. Postoperative AV nodal function was decreased (as measured by the cycle length [CL] at which Wenckebach periodicity occurred) in 15 of 55 studies (27%) of the entire group. There was nearly an equal chance for improvement (24%) in function. This also applied to those patients who had sequential studies. Therefore, this method of assessment for AV nodal function was unreliable, or else the operation did not affect the node significantly. The latter is unlikely in view of late postoperative data. The greatest utility of this test was to determine the capability for AV conduction in certain patients with slow escape rhythms in the absence of surface P waves, and to differentiate complete heart block from AV dissociation when atrial activity was absent. Despite the variability of effects on the sinus and AV nodes in these patients, those who demonstrated depression had a significantly higher incidence of dysrhythmias (80% of patients with sinus nodal depression and 100% of patients with AV nodal depression).