To evaluate coronary flow reserve during cardiac catheterization, intracoronary adenosine and papaverine have been used in the clinical setting. Although papaverine maximizes coronary blood flow, it induces several toxic side effects that reduce its desirability as a coronary dilator. This investigation was designed to compare the subselective intracoronary administration of papaverine with that of adenosine in an animal model. In dogs (n = 34), we studied the effects of each agent on hemodynamics, regional myocardial blood flow, contractility (sonomicrometric and echocardiographic), metabolism (coronary arterial and venous lactate and tissue high-energy phosphates), and electrocardiographic (ST and QT intervals) parameters. Barbiturate and morphine anesthesia/analgesia was induced, and a left thoracotomy was performed. An arterial shunt was created from the left carotid artery to the left anterior descending coronary artery. Two separate groups were studied: group 1 (n = 16) for regional myocardial blood flow and mechanical function and group 2 (n = 18) for biochemical measurements. Adenosine (67 +/- 2 micrograms/min) or papaverine (6 +/- 1 mg/min) was infused into the coronary shunt at a rate of 0.5 + 0.1 ml/min for a maximum duration of 3.5 minutes. Regional myocardial blood flows were determined at control (predrug) and maximal coronary flow using radiolabeled microspheres. All hemodynamic, wall motion, biochemical, and electrocardiographic parameters were also measured at these times. Both drugs produced comparable increases in total and regional coronary blood flows (adenosine, 1.21 +/- 0.15 to 4.83 +/- 0.36 ml/min/g; papaverine, 1.21 +/- 0.05 to 4.89 +/- 0.28 ml/min/g) upon infusion into the left anterior descending coronary artery. Papaverine produced significant (p less than 0.05) changes in subendocardial ST segment electrocardiogram (-2.5 mm), QT prolongation (8 +/- 2%), myocardial creatine phosphate (47% decrease), and coronary sinus serum lactate (277% increase) compared with control. In addition, intracoronary papaverine induced an abnormal contractile pattern. No significant changes in any of these parameters (i.e., ST segment, QT prolongation, myocardial creatine phosphate level, or lactate level) were observed with intracoronary adenosine infusions. We conclude that intracoronary adenosine is comparable to papaverine for maximizing coronary blood flow without the deleterious properties observed with intracoronary papaverine.