The purpose of this study was to determine the clinical feasibility of diagnosing significant coronary artery disease by positron imaging of myocardial perfusion without a cyclotron, using generator-produced rubidium-82 (82Rb). Fifty patients underwent positron emission tomography of the entire heart using a multislice positron camera and intravenous 82Rb or nitrogen-13 ammonia (13NH3) before and after intravenous dipyridamole combined with handgrip stress. Images were read by two observers blinded as to clinical or arteriographic data. Automated quantitative coronary arteriography was obtained for the arteriographic determination of coronary flow reserve, previously demonstrated to be a single integrated measure of stenosis severity accounting for all its geometric dimensions of length, absolute diameter, percent narrowing and asymmetry by quantitative analysis of cine films. Significant coronary artery disease was defined as an arteriographically determined coronary flow reserve of less than 3.0 based on all stenosis dimensions. Any single geometric measure of stenosis severity alone was an inadequate reference standard for comparison with perfusion images. Sensitivity of identifying patients with coronary artery disease having an arteriographically determined coronary flow reserve of less than 3.0 was 95% by positron imaging with a specificity of 100%. The single case that was missed, studied with 13NH3, had a 43% diameter narrowing of a small ramus intermedius off the left coronary artery with no significant narrowing of the major coronary arteries. Positron emission tomography of myocardial perfusion before and after intravenous dipyridamole combined with handgrip stress utilizing generator-produced 82Rb provides sensitive and specific diagnosis of reduced coronary flow reserve due to coronary artery disease in humans.