The control of the patency of aortocoronary bypass grafts necessitates further coronary angiography, an invasive investigation which is difficult to perform routinely because over 60 p. cent of operated patients are asymptomatic. Non-invasive methods have been proposed for this task, including computerised axial tomography (CAT) and exercise Thallium 201 myocardial scintigraphy (EMS). The aim of this study was to assess the relative value of CAT and EMS, alone and in association, in comparison with coronary angiography. Thirty six patients (35 men, 1 woman) with a mean age of 54 years were studied. These patients had a total of 59 bypass grafts inserted an average of 23 months before investigation (20 single, 10 double, 5 triple and 1 quadruple bypass grafts). CAT scanning was performed the day before coronary angiography. Sections of the thorax 7 mm thick were recorded after intravenous injection of contrast medium. A patent graft was identified as an opacity increasing after the injection of contrast on one of the aortic walls. During coronary angiography a graft was declared to be patent when it was opacified selectively or during aortography, and when the grafted coronary artery was seen to be revascularised. The quality of the distal coronary bed was also evaluated (implantation of the graft, distal and collateral vessels). Twenty three patients (with a total of 36 grafts) also underwent EMS on the same day as CAT scanning. Normal fixation in the revascularised territory was taken as evidence of a patent graft. --Coronary angiography showed that 44/59 grafts were patent at 24 months. --CAT scanning was unable to evaluate 12/59 grafts. Thoracic metallic clips created stratified artifacts and analysis of the section was impossible (20 p. cent of CAT investigations were non-contributive); of the interpretable investigations, 40/47 grafts were correctly assessed (85 p. cent): 30/32 patent grafts and 10/15 occluded grafts. --The results of EMS were less reliable; 23/36 grafts correctly assessed (64 p. cent), 18/27 patent grafts and 5/9 occluded grafts. However, EMS provides complementary information to that provided by CAT scanning, especially with respect to the distal coronary bed. When the two methods were used together, 15/15 good surgical results (patent grafts with good distal vascularisation) and 6/8 poor results (patent grafts but poor distal vascularisation), were identified. We conclude that these two atraumatic methods, CAT scanning and EMS, which may be performed on out-patients, are valuable for the routine assessment of the patency of coronary bypass grafts. Coronary angiography could therefore be reserved for those patients in whom further surgery is being considered.