Background: Basic life support (BLS) guidelines focus on chest compressions with a minimal no-flow fraction (NFF), early defibrillation, and a short perishock pause. By using an automated external defibrillator (AED) lay persons are guided through the process of attaching electrodes and initiating defibrillation. It is unclear, however, to what extent the voice instructions given by the AED might influence the quality of initial resuscitation. Methods: Using a patient simulator, 8 different commercially available AEDs were evaluated within two different BLS scenarios (ventricular fibrillation vs. asystole). A BLS certified instructor acted according to the current European Resuscitation Council 2010 Guidelines and followed all of the AED voice prompts. In a second set of scenarios, the rescuer anticipated the appropriate actions and started already before the AED stopped speaking. A BLS scenario without AED served as the control. All scenarios were run three times. Results: The time until the first chest compression was 25 ± 2 seconds without the AED and ranged from 50 ± 3 to 148 ± 13 seconds with the AED depending on the model used. The NFF was .26 ± .01 without the AED and between .37 ± .01 and .72 ± .01 when an AED was used. The perishock pause ranged from 12 ± 0 to 46 ± 0seconds. The optimized sequence of actions reduced the NFF, which ranged now from .32 ± .01 to .41 ± .01, and the perishock pause ranging from 1 ± 1 to 19 ± 1seconds. Conclusions: Voice prompts given by commercially available AED merely meet the requirements of current evidence in basic life support. Furthermore, there is a significant difference between devices with regard to time until the first chest compression, perishock pause, no-flow fraction and other objective measures of the quality of BLS. However, the BLS quality may be improved with optimized handling of the AED. Thus, rescuers should be trained on the respective AED devices, and manufacturers should expend more effort in improving user guidance to shorten the NFF and perishock pause.