Cardiac resynchronization therapy (CRT) is a new and promising therapeutic option for patients with severe heart failure and intraventricular conduction delay. Patients who are candidates for CRT and have a previously implanted device may utilize a "Y" IS 1 connector to accommodate the coronary sinus lead. This modification has the potential to alter biventricular pacing thresholds. During an 18 month period, successful biventricular pacemaker implantation was performed in 72 patients (age: 67 +/- 11 years, left ventricular ejection fraction: 20.5 +/- 5.6%). All of these patients had severe symptomatic congestive heart failure (NYHA Class III and IV). In 20 patients a special "Y" adaptor that bifurcates the ventricular IS 1 bipolar output to two bipolar outputs or one unipolar and one bipolar output was utilized. During initial implantation, LV thresholds obtained in a unipolar configuration prior to connecting to the "Y" adaptor were significantly lower than thresholds obtained after connecting to the "Y" adaptor (1.7 +/- 1.11 V at 0.5 ms pulse width versus 2.8 +/- 1.5 V at 0.5 ms pulse width [P = 0.01]). Two patients (10%) required left ventricular lead revisions due to unacceptably high left ventricular thresholds during device follow-up. The difference in measured left ventricular thresholds between the two configurations is best explained by a resistive element that is added to the circuit when performing threshold measurement of the LV lead through the "Y" adaptor (combined tip to RV ring configuration) versus measurement of the LV lead in a unipolar configuration. This resistive element represents multiple factors including anode surface area, resistive polarization at the tissue-electrode interface, and transmyocardial resistance. LV thresholds should be measured in an LV tip to RV ring configuration or ideally in a combined tip (LV and RV) to shared ring configuration in order to accurately assess LV thresholds. This observation has significant clinical implications as loss of capture may occur as a result of improper measurement of left ventricular thresholds at the time of implantation.