Introduction: The semi-automatic external defibrillator (AED) demands interruptions in chest compression for rhythm analysis, charging and shock delivery in the management of cardiac arrested victims. The introduction of AEDs has enabled defibrillation by laymen, which shortens the interval from collapse to defibrillation. Clinical studies, however, have shown conflicting results as to the effect on long-term survival. A possible explanation to this might be that AEDs waste time that might otherwise be used for CPR. We hypothesised that time without chest compressions during resuscitation by ambulance personnel would be shorter when the defibrillator was operated in manual mode compared to semi-automatic mode.Materials and methods: Ambulance personnel performed CPR on a manikin with the defibrillator in semi-automatic (AED) or manual mode (MED) Time spent on chest compression, rhythm analysis and shock delivery were recorded.Results: Median time from last chest compression to delivery of first shock was 17s for AED vs. 12s for MED (p=0.004). Median time from last shock to resumed chest compression was 25s for AED vs. 8s for MED (p=0.001). Total proportion of time without chest compressions (the hands-off-ratio) was 0.3 for AED vs. 0.2 for MED (p=0.001). Sensitivity for VF detection was 100% in both groups, with a specificity of 100% and 89% for AED and MED, respectivelyConclusion: Hands-off-ratio was significantly higher when ambulance personnel performed CPR with the defibrillator in semi-automatic mode compared to manual mode.