Non-selective intra-arterial digital subtraction angiography (DSA) was performed immediately before selective coronary and bypass angiography in 33 consecutive symptomatic patients 48 +/- 30 months after coronary surgery, for the assessment of 75 coronary bypass grafts. Forty ml of non-ionic, low-iodine content contrast medium (iohexol) were injected into the ascending aorta at 10-20 ml/sec through a 7 or 8 F femoral pigtail catheter. Electrocardiogram-triggered images were acquired on a Siemens Digitron II apparatus in multiple projections in 24 patients and in a single projection in 9 patients. The results of this technique were compared by two independent angiographers with those of selective graft angiography in the same patients. Patency was shown by DSA in 45 of 54 grafts confirmed to be open by selective angiography (sensitivity 83%). Of 21 occluded grafts, stumps were clearly visible at selective angiography in 18 and at DSA in 9 (sensitivity for graft stumps = 50%, p less than 0.01). Of 54 patent grafts with selective angiography, the distal anastomosis could be visualized by DSA in 28 (52%), but the resolution was comparable to selective angiography in 20 grafts (37%) only. A non-significant difference in the sensitivity of DSA was observed between patent saphenous grafts to the left anterior descending coronary artery versus all other coronary arteries (95 vs 85%, respectively), while only 1 of 5 patent left internal mammary artery grafts to the left anterior descending coronary artery was visualized. In 16 of 50 grafts (32%) visualized in a second projection substantial additional diagnostic information was obtained. In conclusion, non-selective intra-arterial electrocardiogram-triggered DSA can visualize patent saphenous grafts with a high sensitivity and may be a useful screening tool for bypass grafts patency; false negatives, however, and poor visualization of distal anastomoses limit its routine clinical use.