BACKGROUND: Among the various physiochemical stimuli, hypoxia has been known to cause coronary vasodilation. In contrast to this, endothelial dependent contracting factor(EDCF) was shown to be secreted by hypoxia and overall physiological roles of these apparently contradicting two phenomena are not clear. Although coronary vasodilation is dominant in epicardial coronary artery by hypoxia, collateral circulation may show different response from epicardial coronary artery to the same stimulus and effect of hypoxia on the vasomotor tone of collateral vessels has not been established. METHODS: Left circumflex coronary artery was chronically occluded using Ameriod occluder in the canine model and myocardial blood flow through collateral circulation was measured using microsphere during induced regional hypoxia. RESULTS: 1) Myocardial blood flow measurements during oxygenated and hypoxic solution infusion were 1.11+/-0.11 mg/min/g and 1.12+/-0.10 ml/min/g respectively in normal perfused zone(LAD territory), but in the collateral dependent zone(LCX territory) blood flow decreased significantly during hypoxic solution infusion(0.55+/-0.17 ml/min/g vs 0.43+/-0.21 ml/min/g)(p<0.05). Also myocardial blood flow ratio(LCX/LAD territory) decreased significantly during hypoxic solution infusion(0.49+/-0.16 vs 0.39+/-0.02)(p<0.05). 2) In collateral dependent zone, endocardial and epicardial blood flow ratio showed significant redistribution during hypoxic solution infusion. 3) After verapamil administration, myocardial blood flow in collateral dependent zone increased from 0.43+/-0.21ml/mg/g to 0.56+/-0.23 ml/mg/g(p<0.05). Also myocardial blood flow ratio(LCX/LAD territory) increased from 0.39+/-0.20 to 0.50+/-0.20 to 0.50+/-0.21 after verapamil administration. CONCLUSION: Hypoxia seems to cause vasoconstriction in collateral vessels and redistribution of blood flow in collateral dependent zone and these effects can be reversed by verapamil.