An ideal coronary vasodilator for studying coronary flow reserve in humans would rapidly produce maximal coronary vasodilation, be short acting to permit repeated measurements, and not alter systemic hemodynamics. The two commonly used vasodilators (dipyridamole and meglumine diatrizoate) do not satisfy these criteria; meglumine diatrizoate does not produce maximal hyperemia and dipyridamole has a long duration of effect (greater than 30 min). In this study we used a subselective coronary Doppler catheter to measure the dose-response kinetics of a shorter acting vasodilator, intracoronary papaverine. In 10 patients with normal coronary vessels, the maximal vasodilator response to papaverine was compared with that to intravenous dipyridamole (0.56 mg/kg infused over 4 min) and intracoronary meglimine diatrizoate. The increase in coronary blood flow velocity after the maximal dose of papaverine (4.8 ± 0.4 peak/resting velocity ratio, mean ± SEM) was nearly identical to that seen after infusion of dipyridamole (4.8 ± 0.6) and was significantly greater than that after meglumine diatrizoate (3.1 ± 0.2, p < .01). At maximal hyperemia, mean arterial blood pressure fell 9 ± 2% (mean ± SEM) after intracoronary papaverine, 8 ± 4% after dipyridamole, and 3 ± 3% after meglumine diatrizoate. The dose-response kinetics of intracoronary papaverine were studied in 13 patients with normal coronary arteries. In the left coronary artery, maximal vasodilation (5.4 ± 0.6) was achieved with 8 mg in six of eight patients and with 12 mg in all patients. In the right coronary artery, maximal vasodilation (4.8 ± 0.7) was achieved with 6 mg in four or five patients and with 8 mg in all patients. Onset of maximal vasodilation was rapid after papaverine (16 ± 1 sec) and meglumine diatrizoate (15 ± 1), but prolonged after dipyridamole (4.8 ± 0.4 min after onset of infusion). The duration of maximal vasodilation was brief after papaverine (49 ± 10 sec) and meglumine diatrizoate (8 ± 1 sec), but prolonged after dipyridamole (greater than 4 min). Coronary blood flow velocity returned to within 10% of resting values quickly after papaverine (128 ± 15 sec) and meglumine diatrizoate (42 ± 4 sec). After dipyridamole infusion, however, coronary blood flow velocity remained elevated for greater than 4 min after completion of infusion. These results suggest that papaverine can produce intense, rapid-acting vasodilation of the coronary arteriolar bed equivalent to that stimulated by intravenous dipyridamole without markedly altering systemic arterial pressure. Although the extent and duration of vasodilation was dose dependent, the hyperemic period in all patients was sufficiently brief to allow multiple measurements of coronary reserve over a short period of time. The use of intracoronary papaverine to measure the maximal flow reserve capacity of individual coronary vessels should greatly facilitate studies of the coronary circulation in patients undergoing cardiac catheterization.