Objective To investigate the efficiency and safety of the automated external defibrillator (AED); and to compare the defibrillating efficacy between the domestic AED and the imported AED. Method Fourteen Peking white swine weighed (30 ± 1 ) kg, were anaesthetized and arterial cannula was inserted into the left femoral artery for collecting blood samples and measuring artery blood pressure, cardiac output (CO) and external vascular lung water index (EVWI) by using PiCCO monitor, and another cannula was inserted into the left femoral vein to place an electrode as a temporary pace maker, and then ventricular fibrillation (VF) was induced by using a programmed electrical stimulation instrument. The swine were divided into two groups randomly ( random number) after VF was confirmed by electrocardio-monitor,and the domestic made AED (M group, n = 7) or imported AED (Z group, n = 7) were used respectively in 7 swine of each group. The electrodes of defibrillator were stuck to the precordium and sternum separately, and defibrillation was delivered by the AED signal. The entire procedure of defibrillation was repeated four times. The number of defibrillation delivery required to get normal cardiac rhythm was recorded. Twenty minutes after restoration of spontaneous circulation (ROSC), blood sample was taken to measure myocardial enzyme. The CO and EVWI were measured too. Measurement data was analyzed by repeated ANOVA and enumeration data was analyzed by Chi-square test. A value of P ＜0.05 was regarded as being significant statistical difference. Results There were 54 episodes of VF induced in 14 swine and only one swine in Z group was dead after 2nd episode of VF, resulting in 98.1% success rate of resuscitation. There were 37 defibrillation deliveries given to swine of M group, and 75.0% produced normal cardiac rhythm after the first delivery of defibrillation,whereas 80.0% in Z group after 32 defibrillation deliveries. The length of time from AED turned on to the recognition of ECG signal of VF was (29 ± 1 ) s. There were 2 occasions in both groups that the AEDs failed to recognize the VF signal emerged on ECG after the first induction of VF by electrical shock. Similarly, there were 2 occasions in Z group that the AED incorrectly recognized the signal of ventricular tachycardia as that of VF, but no harm was done after defibrillation delivery given following the signal of AED. During the entire course of experiment, heart rate,blood pressure and CO of all survived swine were stable,but the EVWI and myoglobin (MYO) gradually increased as time elapsed until the 3rd or 4th episode of VF leading to significant differences from the original ones. There were no significant differences in all biomarkers between two groups. Conclusions The VF can be ceased by AED safely and efficiently. There are no significant differences in signal recognition, efficacy of defibrillation and myocardial injury found between the domestic made AED and imported AED; but AED may not be good to be recommended to the professional staff of resuscitation because of its incorrect reorganization of ECG owing to over-automation.