Mainly because of technical problems, the use of rabbits as a cardiopulmonary bypass (CPB) animal model with direct cannulation of the ascending aorta is known to be extremely difficult. The objectives of this study were the establishment of a CPB model in rabbits with direct cannulation of the ascending aorta, and the evaluation of the protective effect of steroid on the development of brain edema during circulatory arrest (CA) in an established rabbit CPB model. Fifteen New Zealand white rabbits were divided into three groups; control CA group, CA with Trendelenberg position, and CA with Trendelenberg position and steroid administration. After anesthetic induction and tracheostomy, median sternotomy was performed. An aortic cannula (3.3 mm) and a venous cannula (14 Fr) were inserted into the ascending aorta and the right atrium, respectively. The CPB circuit consisted of a roller pump and a bubble oxygenator. With 120-150 ml of blood, the priming volume of the circuit was approximately 450 ml, and CPB at a flow rate of 80-85 ml/kg/min was initiated. Blood in the priming solution was obtained from donor rabbits through cardiac puncture. Ten minutes later, CA with cessation of CPB was established for 40 min at 20 degrees C (rectal temperature). After CA, CPB was restarted with a 20 min period of rewarming. Ten minutes after weaning, the animal was sacrificed. Between 1 and 2 g of the brain was removed and the water content was determined and compared between groups. CPB with CA was successfully performed in all cases, with a flow rate of 60-100 ml/kg/min maintained throughout the CPB procedure. At that time, blood gases were reasonably maintained and aortic pressure ranged from 35 to 55 mmHg. After weaning from CPB, all hearts resumed beating spontaneously. Among the three groups, there were no statistically significant differences in the water content of the brain. These results indicate that: (1) if the proper technique is used, CPB in rabbits with direct cannulation of the ascending aorta is a reliable procedure, and (2) the effect of steroid on the prevention of brain edema related to the Trendelenburg position during CA is not established within the scope of this study.