The hypothesis that changes in myocardial interstitial fluid (ISF) adenosine (Ado) concentrations couple coronary blood flow (CBF) with myocardial oxygen need is still controversial. Measurements of Ado in cardiac tissue, coronary sinus plasma, and pericardial infusates have commonly been used as indexes of ISF Ado, but recent studies have highlighted the potential invalidity of many of the assumptions underlying these sampling methods. With the use of an epicardial chamber (EC) to contain a small volume of buffer on the undisturbed left ventricular epicardium of anesthetized open-chest dogs, estimates of ISF Ado can be obtained without invoking these assumptions. In experiments designed to validate our method and investigate the role of Ado in the regulation of CBF during myocardial catecholamine stimulation, we found that endogenous Ado accumulated in the EC exponentially and attained constant steady-state concentrations, ranging from 75 to 125 nM, within 4 min. Intravenous infusions of two catecholamines increased steady-state EC Ado concentrations twofold, but did not affect Ado influx kinetics. We conclude that the EC technique yields more direct estimates of steady-state ISF Ado concentrations than previously available and that ISF Ado is involved in the mediation of the hyperemia accompanying myocardial catecholamine stimulation.