OBJECTIVES We sought to ascertain the utility of adenosine thallium-201 tomography for assessing graft stenoses late after coronary artery bypass graft surgery. BACKGROUND Although pharmacologic perfusion imaging has been increasingly used in the assessment of patients with coronary artery disease, the value of this stress modality for detecting coronary artery bypass graft stenosis late after surgery is unknown. METHODS We studied 109 patients who underwent both adenosine thallium-201 tomography and coronary angiography at 6.7 +/- 4.8 (mean +/- SD) years after coronary artery bypass graft surgery. Adenosine thallium-201 tomography was assessed quantitatively by computer-generated polar maps of the myocardial thallium-201 activity. RESULTS On coronary angiography, significant graft stenoses were present in 68 patients, 65 of whom had a corresponding perfusion defect as shown by thallium-201 tomography (sensitivity 96%). Significant stenoses were present in 107 (37.8%) of 283 grafts. The overall specificity by quantitative tomography was 61%. Seventy percent of the apparently false positive perfusion defects could be explained on the basis of unbypassed native disease or by the presence of fixed defects in patients with previous myocardial infarction. CONCLUSIONS Thus, results of adenosine thallium-201 tomography are nearly always abnormal in patients with late coronary graft stenosis. Most of the false positive defects appear to be due to either unbypassed native disease or a previous myocardial infarction.