An improved assay was used to investigate the effects of hypoxia or ischemia on interstitial fluid and coronary venous effluent levels of adenosine in isolated perfused nonworking rat hearts. The adenosine in 5- to 10-μl samples of left ventricular epicardial surface transudates and coronary effluents was reacted with chloroacetaldehyde, and the fluorescent derivative (l, A6-ethenoadenosine) was quantitated using high pressure liquid chromatography and fluorescence detection. Hearts responding to hypoxia could be separated into two groups. In one group of hearts, the control (normoxic) transudate and effluent adenosine concentrations were 94 ± 24 and 41 ± 6 pmol/ml, respectively. These values increased by 118 and 96%, respectively, with 5 minutes of hypoxia (30% O2, and returned to control levels 5 minutes after resumption of normoxia. In a second group of hearts, the normoxic control levels of adenosine in the transudates (42 ± 7 pmol/ml) and coronary effluents (62 ± 17 pmol/ml) were increased with hypoxia by 174 and 1,178%, respectively. However, the transudate levels continued to rise for 5 minutes after resumption of normoxic perfusion while effluent levels fell. In another series of hearts, global ischemia for 30 seconds elicited an elevation of transudate adenosine levels by 362 to 641 % above control (58 ± 15 pmol/ml) as determined 30 seconds after resumption of perfusion flow. These data suggest that 1) the adenosine concentration in the interstitium of the oxygenated nonworking rat heart is 4 × 10−8to 1 × 10−7M, and 2) coronary effluent levels of adenosine neither accurately reflect the absolute magnitudes of interstitial adenosine to which myocardial cells are exposed nor consistently parallel changes in interstitial adenosine levels that may occur with alterations in oxygen delivery.