To assess the effect of regular and high-dose dipyridamole on coronary flow velocity in the left anterior descending artery (LAD), and to determine whether assessment of coronary flow velocity reserve (CFVR) is more sensitive for detection of ischemia than standard echocardiographic criteria, 47 patients were studied prospectively: 16 patients with stenosis of the LAD, 18 patients with angiographically normal LADs, and 13 patients with minimal disease. Patients underwent transesophageal echocardiographic study of wall motion and LAD flow velocity at baseline and at hyperemia, and for angina and electrocardiographic changes. The mean CFVR values after 0.56 mg/kg after 0.84 mg/kg of dipyridamole were similar: 2.52 +/- 0.87 versus 2.62 +/- 0.90. A CFVR <2.3 (normals mean -2 SDs) was more sensitive (88% at both doses) for the detection of underlying coronary obstruction than was wall motion monitoring (44% and 75%, respectively). The combination of CFVR <2.3 and wall monitoring was more sensitive than index alone (94% at both 0.56 and 0.84 mg/kg). The rate-pressure product was not significantly different at the two doses of dipyridamole. When flow response is the end point of stress testing, as with transesophageal monitoring, the 0.56 mg/kg dose of dipyrid mole is adequate, but when ischemia is the end point (as with wall motion monitoring by 2-dimensional echocardiography), the dose of 0.84 mg/kg is more sensitive.