The purpose of this study was to define the reproducibility of sequential quantitative exercise thallium-201 tomography. This was an ancillary study of a randomized, double-blind, placebo-controlled trial evaluating the short-term efficacy of transdermal nitroglycerin patches in stable patients with angiographic coronary artery disease and no prior myocardial infarction. All 18 patients had a baseline tomographic perfusion defect involving > or = 5% of the left ventricle after treadmill exercise. At a minimum of 3 days (mean 6.1 +/- 1.8) after double-blind randomization to placebo, exercise thallium-201 tomography was repeated (study 2) using the same exercise protocol as in the baseline study (study 1). No significant differences in exercise parameters were observed from studies 1 to 2. Seventeen of 18 patients (94%) had an abnormal repeat exercise perfusion scan and 96% of initially abnormal vascular territories remained abnormal. The mean tomographic perfusion defect size was not significantly different from studies 1 (17.4 +/- 13.3%) to 2 (16.6 +/- 15.3%), nor were the components defined as scar and ischemia. A > or = 10% change in total perfusion defect size in an individual patient defined the 95% confidence interval for exceeding the variability of the tomographic technique. Quantitative exercise thallium-201 tomography is highly reproducible and can be used to accurately interpret temporal changes in myocardial perfusion in individual patients.