Implantable cardioverter defibrillators (ICDs) with dual chamber and dual chamber rate responsive pacing may offer hemodynamic advantages for some ICD patients. Separate ICDs and DDDR pacemakers can result in device to device interactions, inappropriate shocks, and underdetection of ventricular fibrillation (VF). The objectives of this study were to compare the VF detection times between the Ventak AV II DR and the Ventak AV during high rate DDDR and DDD pacing and to test the safety of dynamic ventricular refractory period shortening. Patients receiving an ICD were randomized in a paired comparison to pacing at 150 beats/min (DDD pacing) or 175 beats/min (DDDR pacing) during ICD threshold testing to create a "worst case scenario" for VF detection. The VF detection rate was set to 180 beats/min, and VF was induced during high rate pacing with alternating current. The device was then allowed to detect and treat VF. The induction was repeated for each patient at each programmed setting so that all patients were tested at both programmed settings. Paired analysis was performed. Patient characteristics were a mean age of 69 +/- 11 years, 78% were men, coronary artery disease was present in 85%, and a mean left ventricular ejection fraction of 0.34 +/- 0.11. Fifty-two episodes of VF were induced in 26 patients. Despite the high pacing rate, all VF episodes were appropriately detected. The mean VF detection time was 2.4 +/- 1.0 seconds during DDD pacing and 2.9 +/- 1.9 seconds during DDDR pacing (P = NS). DDD and DDDR programming resulted in appropriate detection of all episodes of VF with similar detection times despite the "worst case scenario" tested. Delays in detection may be seen with long programmed ventricular refractory periods which shorten the VF sensing window and may be avoided with dynamic ventricular refractory period shortening.