This report describes a cost analysis for clinical positron emission tomography (PET) of the heart using generator produced rubidium-82 (82Rb). Considered sequentially are the clinical problem, current noninvasive radionuclide methods, positron emission tomograph, and the cost of PET per study. Also analyzed are the costs of PET versus thallium imaging in the management of chest pain, for screening asymptomatic men at high risk for coronary artery disease and for evaluating myocardial viability after myocardial infarction or thrombolytic therapy. Noninvasive assessment of coronary artery stenosis and myocardial ischemia/viability in symptomatic or asymptomatic subjects remains a major medical problem because the sensitivity and specificity of thallium imaging are only 70-85% and 50-70%, respectively, in recent studies. Cardiac positron imaging has an accuracy for noninvasive diagnosis of coronary artery disease in symptomatic or asymptomatic patients with a sensitivity and specificity of 95-98%. It can also be used for assessing physiologic stenosis severity, for imaging myocardial infarction and viability, for assessing effects of interventions such as thrombolysis, percutaneous transluminal coronary angioplasty (PTCA) or bypass surgery on myocardial perfusion, metabolism or coronary flow reserve, for assessing collateral function noninvasively in man, and for diagnosing cardiomyopathy not due to coronary artery disease. Although the cost for cardiac PET with 82Rb may be modestly higher than for 201Tl, the greater diagnostic yield of PET results in comparable or lower overall medical management costs than no diagnostic tests/interventions and lower overall costs compared to thallium imaging for evaluating patients with chest pain, asymptomatic high risk males, and patients after acute myocardial infarction/thrombolysis for myocardial viability.