A prospective study of 125 femoro-distal vein bypass grafts (103 patients) was undertaken to assess the role of treadmill exercise testing in the detection and correction of early vein graft stenosis. Patients were followed at 1,3,6,12 months and then annually with routine clinical assessment, palpation of pulses and resting Doppler ankle/brachial index. In addition, the Doppler index was repeated after 5 minutes exercise on a treadmill set at 5 degrees and 3 km/hr. Follow-up has been carried out for a mean of 24 months (range 3-48 months) and during that time 30 vein grafts (28 patients) were submitted to angiography. This revealed vein graft stenosis in 15, native vessel disease in 11 and no morphologic abnormality in 4. One additional graft was suspected to have a stenosis but occluded before angiography could be carried out. At the time of detection by the treadmill test, 8 of the 15 (53%) confirmed vein graft stenoses, and 4 of 11 (36%) native vessel stenoses would not have been detected by clinical assessment or resting Doppler pressure measurements. The 15 vein graft stenoses which were corrected all remain patent, suggesting that treadmill testing is of value in the detection and prevention of early femoro-distal vein graft occlusion.