Nonthoracotomy lead systems for implantable cardioverter defibrillators (ICDs) have reduced operative mortality and morbidity as compared to epicardial lead systems but are usually associated with higher defibrillation thresholds (DFTs). The purpose of this prospective randomized trial was to investigate if the second defibrillation electrode in the left subclavian vein can increase defibrillation efficacy and decrease DFT as compared to the superior vena cava (SVC) position in nonthoracotomy lead systems for ICDs. Seventeen patients (mean age: 49.9 +/- 11.3 years, mean ejection fraction: 46.1% +/- 15.8%) were implanted with an investigational unipolar electrode (Medtronic 13001) used as the defibrillation anode. DFT testing was started in the SVC (n = 10, group A) or the left subclavian vein (n = 7, group B), and repeated in the alternative position starting at the DFT of the initial position. Fifteen patients were eligible for analysis (group A: n = 9, group B: n = 6). With the electrode in the SVC, ventricular fibrillation could be successfully terminated in 9 out of 15 patients (60%). In the left subclavian vein the success rate was 100% (P < 0.01). Mean DFT in the SVC was 13.0 +/- 5.2 J and in the left subclavian vein 10.2 +/- 4.9 J. DFTs in the left subclavian vein were either lower (group A: n = 5/9, group B: n = 5/6) or equal to the results in the SVC position (P < 0.001). Thus, the left subclavian vein appears to be a superior alternative for positioning of the defibrillation anode as compared to the SVC for nonthoracotomy lead systems using two separate leads.