A 70-year-old man with coronary artery disease, chronic atrial firillation, and cardiomegaly was scheduled for redo coronary bypass urgery. His initial heart rate was 38 beats/min. Because of the bradyardia and the epicardial scarring expected from the previous surgery, pacing probe was made ready in the right ventricular channel of a aceport pulmonary artery catheter (Edwards Lifesciences, Irvine, CA) nserted via the right internal jugular vein.4,5 After surgical revascularzation (performed with cardiopulmonary bypass and hypothermic poassium cardioplegia), the patient underwent electrical defibrillation, hich restored sinus rhythm with first-degree atrioventricular block. he surgeon attached pairs of epicardial pacing wires to the right trium and the right ventricle. After chest closure, the patient’s rhythm everted to atrial fibrillation. Subsequently, the thresholds for the epiardial ventricular pacing wires increased and then failed to capture at he maximal output (25 mA) of the temporary pacemaker (Medtronic 388, Minneapolis, MN). The endocardial ventricular pacing probe was dvanced 4 cm from its channel, and transvenous pacing was successully instituted with the Medtronic 5388 pacemaker programmed to eliver 15 mA ventricular pulses at 80 beats/min in the demand mode VVI). Interestingly, both demand (VVI) and asynchronous (VOO) transenous ventricular pacing failed whenever the epicardial atrial wires ere detached from the pulse generator. Even though the atrial ortion of the dual-chamber pacer was turned “off,” the atrial ystem participated in ventricular pacing (Fig 1). The circuit is iagrammed in Figure 2. The epicardial atrial wires were no longer required for ventricular acing when the endocardial ventricular pacing probe was advanced to rotrude 6 cm, instead of 4 cm, from the Paceport catheter. The aceport probe is comprised of 2 wires, one of which projects farther han the other. Advancing the probe enabled both wires to make contact ith the right ventricular endocardium. With both Paceport wires in ontact with the right ventricular endocardium, attached atrial wires ere not required for ventricular pacing. However, when only the distal ire of the Paceport probe made contact with the ventricular endocarium, then an effective ventricular pacing circuit was completed by eans of the epicardial atrial wires and the atrial portion of the ual-chamber pacer. By means of an oscilloscope (Tektronix digital model 2430, Beaerton, OR), this unconventional pacing possibility was duplicated in he laboratory. A 500-ohm resistor was run from the V pole of the ulse generator to either the V pole or the A pole. The pacemaker as set to deliver 15 mA pulses at 80/min in the VVI mode, and the otential difference across the resistor was tracked in time. Four edtronic 5388 pacemakers were examined. As expected, the V V rrangement produced strong electrical pulses. The amplitudes were .62 0.12 V, and the widths were 1.52 0.02 milliseconds. Howver, the V A arrangement also produced significant pulses (0.58 .04 V and 1.51 0.04 milliseconds). Significant pulses were also etected in a V A arrangement at higher amperage, but significant ulses were not detected in V A or in V A situations. p