The bidirectional cavopulmonary shunt may be useful as an intermediate procedure before Fontal oper- ation in high-risk patients, because it provides adequate relief of cyanosis and relief of ventricular volume overload. But there are no established theory about the effects of bidirectional cavopulmonary shunt on pulmonary arterial development. The purpose of this article is a study of changes of pulmonary artery size after bidirectional cavopulmonary shunt. Ca diac catheterization and angiography procedures were done on 19 patients who underwent bidirectional cavopulmonary shunt from February 1992 to July 1994, their results were reviewed. Preoperative cardiac catheterization and angiography procedures were performed at a mean interval of ).8 $\pm$4.8($\pm$SEM) months before surgery and following catheterization at a mean postoperative interval of 19.6 $\pm$4.8 months. Pulmonary arterial sixte measurement were standardized for body surface area(Pulmon- arty artery index), and for diameter of descending thoracic aorta(McGoon ratio). Patient's age, body stir- face area, pulmonary angioplasty, preoperative McGoon ratio and follow-up intervals were considered as variables. Before bidirectional cavopulmonary shunt, patient's mean age, body surface area, arterial 02 saturation, diameter of right pulmonary artery, diameter of left pulmonary artery, pulmonary artery Index, McGoon ratio were 13.7$\pm$15.6 months, 0.40$\pm$0.12m2, 71.4$\pm$12.4m2, 7.1$\pm$1.7mm, 6.2$\pm$1 , 191.8$\pm$82.7mm21m2, 1.73 $\pm$0.49, respectively. After bidirectional cavopulmonary shunt, the values were changed to 39.9 $\pm$ 16.2 months, 0.58$\pm$0.07 m2, 83.0$\pm$3.8m2, 9.0$\pm$ 1.5 mm, 7.7$\pm$2.0, 197.3$\pm$57.1 mm2/m2, 1.76$\pm$0.32, respect- ively With patients'development (age, body surf'ace area), diameters of pulmonary arteries were increased, but pulmonary artery indices and McGoon ratios were not changed. And there were no effects of age, body surf'ace area, amount of increased 02 saturation, pulmonary angiography and follow-up duration on the increment of pulmonary ar- tery size. But when the McGoon ratio was as low as 1.2, there were significant increase in postoperative pulmonary artery sizes. There was a significant correlation between preoperative pulmonary artery index (PAI) and McGoon ratio (MGR) ; PAI : MGRxl18.0-12.4 In conclusion, bidirectional cavopulmonary shunt provides adequate increment of arterial 02 saturation and does not increase the pulmonary artery size. Further investi ation is mandatory to evaluate the effect of pulsatile bidirectional cavopulmonary shunt on pulmonary artery growth.