Objective: To establish a simple

stable and high success rate cardiac arrest (CA) model in rabbits. Methods: Cardiac ventricular fibrillation (VF) was induced in 15 New Zealand rabbits by transoesophageal-chest wall electrode with 35mA alternating current. Five minutes after onset of CA

conventional cardiopulmonary resuscitation (CPR) was initiated. The time spent on inducing cardiac arrest

the cardiac rhythm when the electric stimulation was stopped

the cardiac rhythm when the CPR was started

the time spent on CPR

the defibrillation frequency

the administration frequency

the rate of restoration of spontaneous circulation (ROSC)

the mechanical ventilation time after ROSC

the life time after ROSC and the 72h survival rate of 15 rabbits were recorded respectively. Results: All of 15 rabbits were successfully induced cardiac arrest. The inducing time was 26.47±10.36s. When the electric stimulation was stopped

all cardiac rhythm was ventricular fibrillation (15/15)

no pulseless electrical activity and asystole occurred. During the observation of nonintervention

no automatic cardioversion phenomenon occurred. When the CPR was started

electrocardiogram indicated VF occurred in 73.3% of the rabbits（11/15）and pulseless electrical activity occurred in 26.7% of the rabbits (4/15). The time spent on CPR was 200.07±136.00s. The defibrillation frequency was 1.00±0.76. The frequency of intravenous injection of adrenaline was 2.40±1.12. The rate of ROSC was 93.3% (14/15). The mechanical ventilation time after ROSC was 40.46±17.05min. The death rate within 6h after ROSC was 6.7% (1/15) and the 72h survival rate was 26.6% (4/15). Conclusion: The model of cardiac arrest in rabbits induced by transoesophageal-chest wall electric stimulation is very simple operation

and has a stable result and high success rate. It is a ideal animal model for cardiopulmonary resuscitation study.