OBJECTIVE To investigate different clinical applications of gadolinium-enhanced MR angiography (Gd MRA) using three-dimensional breath-hold GE sequences, without bolus time calculation, in patients with vascular diseases. METHODS Forty-seven patients were examined (49 studies in all). Eleven of them had an abdominal aortic aneurysm, 6 surgical bypass, 7 renal artery stenosis, 3 aortoiliac Wallstent, 7 aortic stent-graft; 11 patients had a suspected condition in the thoracic aorta and pulmonary arteries, 1 had subclavian artery stenosis and 1 suspected axillary artery compression. All patients were submitted to breath-hold Gd MRA, after 30-40 mL Gd, with a 1.5 T magnet (Gyroscan ACS-NT, Philips, The Netherlands) and a standard body coil. The GE sequences were acquired with TR = 8.8, TE = 2.7, FA = 60 degrees, matrix = 163 x 512, with 28 seconds acquisition time. Digital subtraction angiography and intraoperative findings were the reference standards to evaluate the results. RESULTS Thoracopulmonary district: metastatic compression of axillary vessels was found in 1 breast cancer patient; the true and the false lumens and the intimal flap were identified in 2 patients with chronic aortic dissection (Stanford A), and graft patency and complete resolution of the aortic dissection was seen in the patient operated on for acute aortic dissection (Stanford A). The thrombus, lumen, extent and diameter were studied in 2 patients with thoracic aortic aneurysms. Subclavian artery stenosis was demonstrated in 1 patient, which was treated with transluminal angioplasty and stenting. The other 6 patients had normal findings. Abdominal aortic aneurysms (AAA): as for disease extent, breath-hold Gd MRA had 100% sensitivity and specificity compared with surgical findings. Juxtarenal aneurysm extent, which had been missed at DSA, was detected in 1 patient and then confirmed at surgery. Stenosis: comparing DSA and MRA findings in the whole series of patients we had 97.4% agreement (155/159 arteries), that is 76.4% (13/17) arteries) considering only stenoses > 50%. Breath-hold Gd MRA sensitivity and specificity were 100 and 87.5%, respectively, in our 28 stenoses. Bypass, aortoiliac stent, vascular endograft: patency was demonstrated in all the 6 surgical bypass patients, and there was agreement with color Doppler findings in 5 of 6. Breath-hold Gd MRA seems to have no possible applications in the follow-up of percutaneously implanted iliac stents, but we had excellent findings about patency and position of nichel-titanium endografts used for AAA treatment. CONCLUSIONS We optimized breath-hold 3D MRA without bolus transit time calculation and with high-dose Gd in different clinical vascular conditions. In our opinion, 3D GE sequences can replace DSA in selected cases, providing a fast, accurate and noninvasive examination.