To assess the feasibility, safety and diagnostic accuracy of adenosine infusion combined with echocardiography, 73 patients with suspected or known coronary artery disease underwent echocardiography at baseline and during a maximal intravenous adenosine infusion of 140 micrograms/kg per min. Compared with baseline values, adenosine caused an increase in heart rate, a decrease in systolic and diastolic blood pressure and a slight but significant increase in rate-pressure product. The echocardiographic images were digitized and randomly assigned in a quad-screen format for nonbiased interpretation. An ischemic response, defined as a new or worsening wall motion abnormality, developed in 25 patients; a fixed wall motion abnormality was present in 27 and no abnormality in 21. All patients underwent coronary angiography. The sensitivity of adenosine echocardiography for greater than or equal to 75% coronary artery diameter stenosis was 85% (46 of 54), with a specificity of 92% in patients with normal coronary arteries. In the 35 patients with a normal baseline electrocardiogram the sensitivity was 60%; 9 (82%) of 11 patients with multivessel disease were correctly identified. The sensitivity for adenosine electrocardiography (greater than or equal to 1-mm ST depression) was 35% with a specificity of 100%. Side effects were transient and mild; aminophylline was used in two patients. Thus, ischemic changes can be induced in patients with coronary artery disease with intravenous adenosine that, combined with echocardiography, is sensitive for the assessment of ischemic heart disease, particularly in patients with multivessel disease.