My objective was to observe if coronary vasodilator responses after bolus (0.2 ml) intracoronary injection of adenosine (1-5000 /IM) or after coronary artery occlusion (5-30 seconds) are modified over a range from mild to severe hypoxia [outflow oxygen tension (PO2V) = 119 ± 9 to 10 ± 2 (mean ± SEM) torr]. The vasculature of ten isolated, paced (240 beats/min), nonworking guinea pig hearts was perfused by the Langendorff technique with a fortified Krebs-Ringer solution at 37.5°C and at a constant pressure of 55 torr. Graded hypoxia was produced by randomly altering perfusate FO2 to one of five predetermined levels. Individually, maximal hypoxia, maximal adenosine injection, and 30-second occlusion increased vascular conductance by 98, 100, and 98%, respectively. At constant myocardial oxygen consumption (MVO2), PO2V was inversely proportional to baseline conductance (r= 0.42,P< 0.01); during normoxia (PO2v= 119 torr) both log-dose adenosine concentration and occlusion period were directly proportional to peak conductance (r= 0.78,P< 0.01;r= 0.73,P< 0.01). Low doses of adenosine or short occlusions significantly shifted the conductance intercept to higher values but did not change individual PO2V vs. conductance slopes. Similar results were found in the relationship between PO2vand overflow volume after adenosine injection or occlusion. Moreover, both the excess volume flow with low doses of adenosine and the volume of reactive hyperemia were unaffected by PO2V when MVO2 was unchanged. As hypoxia became more severe and MVO2 decreased, reactive hyperemia decreased and the flow debt increased. This study indicates that adenosine or coronary occlusion produces additive, noninteractive coronary vasodilation in the presence of tissue hypoxia. One possible explanation for these results is that as conductance rises, the common receptors for endogenous and exogenous adenosine become increasingly saturated.Circ Res 47: 392-399, 1980