Diagnostic and interventional heart catheterization in peripheral vascular disease often requires due to iliacal disease additional methods of arterial approach besides the Judkin's technique. The percutaneous catheterization of the brachial artery finds widespread use. A major complication linked with this method is an increased rate of thrombotic occlusions at the puncture site. Thus, we investigated in a prospective set-up the ability of duplex ultrasound to identify predictive risk factors for vascular complications. Over a period of 20 months, 8000 patients referred to heart catheterization were studied. Routine catheterization via the femoral route was contraindicated in 34 out of 8000 mostly due to severe peripheral vascular disease with multiple vascular risk factors (diabetes, hypertension, and smoking). 53 patients who had a comparable low risk-profile served as the control group. The brachial artery was examined by ultrasound duplex for vessel anatomy and diameter at the puncture site before coronary angiography. Both groups (patient and control group) showed in 15% a variable anatomy with a premature division of the brachial artery in 6% proximal of the elbow and in 9% already distal to the axillary artery. Because of reduced diameters of these variable vessels no procedure was carried out at these arms. In all cases the opposite arm was successfully used instead, because the variants were always located only at one arm. The diameter of the brachial artery measured in average 5.0 +/- 0.8 mm and 4.8 +/- 0.7 mm in patients and controls, respectively. Women had a significantly smaller vessel diameter than men, measuring a difference of 0.4 and 0.6 mm, respectively (p < 0.05). For coronary angiography 6F and 7F arterial sheats were used equally, and in 32% of all cases a coronary intervention was performed. 31 (91%) procedures were carried out without complications; there was a false aneurysm in 1 patient (3%) and an occlusion of the brachial artery at the puncture site in 2 patients (6%). The occluded vessels of two diabetic women had a reduced diameter at the level of 10% of the standard distribution and an unfavorable ratio of sheat-to-vessel-diameter which lead initially to an obstruction of nearly 50% of the vessel lumen during catheterization. Screening of the brachial artery by ultrasound duplex before a percutaneous catheterization for coronary angiography and intervention showed reproducibly the variable anatomy and differences in vessel diameter, which can be risk factors for thrombotic occlusion. Important details for the location of the puncture site and the possible size of the arterial sheat can be obtained, so that coronary interventions with 7F catheter systems are still practicable. This technique is a simple and efficient method to estimate the relative risk of arterial occlusion prior to percutaneous puncture of the brachial artery, especially in a group of patients with severe atherosclerosis and elevated vascular risk-factors.