The effects of intravenous verapamil (0.15 mg/kg) and propranolol (0.2 mg/kg) with regard to atrioventricular (AV) conduction and tachycardia termination during paroxysmal atrial tachycardia were compared in 2 groups of infants (verapamil n = 14, propranolol n = 18, mean age 80 ± 21 days, range 1 to 364). Using transesophageal recording techniques, tachycardia cycle length, AV intervals and ventriculoatrial intervals were measured before and after drug administration. Both intravenous propranolol and verapamil significantly prolonged tachycardia cycle length and AV interval (cycle length—propranolol 230 ± 30 to 262 ± 33 ms, p <0.05, verapamil 223 ± 38 to 245 ± 32 ms, p <0.05; AV interval—propranolol 98 ± 26 to 126 ± 38 ms, p <0.05, verapamil 96 ± 19 to 109 ± 24 ms, p < 0.05). Neither drug prolonged the ventriculoatrial interval. Tachycardia terminated after intravenous verapamil in 11 of 14 infants (79% efficacy rate). Tachycardia terminated in 0 of 18 after intravenous propranolol (0% efficacy rate). In 8 infants an atrial deflection was recorded on the esophageal electrocardiogram at the time of tachycardia termination after intravenous verapamil, which suggested that tachycardia terminated by block occurring in the AV node. In 2 infants a ventricular deflection was recorded at the time of tachycardia termination after verapamil, which suggested that block occurred in the accessory connection. Both drugs prolonged tachycardia cycle length by prolonging AV conduction to a similar degree. Verapamil terminated tachycardia in most infants whereas propranolol was completely ineffective in terms of tachycardia termination, suggesting that it is not simply prolongation of AV conduction which accounts for the efficacy of verapamil in terminating infant tachycardia. Verapamil can terminate tachycardia by causing block to occur in the accessory AV connection, an effect that has not been observed with propranolol. These data suggest that the effect of verapamil on infant tachycardia is different from that of propranolol.