Background and Objective Elderly patients who have been submitted to coronary bypass grafting with the left internal mammary artery (LIMA) may develop a coronary-subclavian steal syndrome because of a left subclavian artery (LSA) stenosis. Usually stenting of LSA is performed by the standard femoral route with guiding catheter technique, but this technique can be particularly difficult in elderly patients who often have iliac-femoral kinking and aortic tortuosity. We compared a new "ad hoc" brachial artery approach technique with the standard guiding catheter technique through the femoral access. Methods Between January 2005 and September 2006, four patients underwent LSA stenting using the left brachial artery access obtained with a 6F or 7F 45-cm-long valved anti-kinking sheath as the Super Arrow Flex sheath (Arrow International, PA, USA). The sheath was positioned just before the LIMA graft ostium and a 0.035 inch 260-cm-long Storq guidewire (Cordis Inc., Johnson & Johnson, Warren, NJ) was advanced across the lesion to the descending aorta. A balloon-expandable Genesis (Cordis Inc., Johnson & Johnson, Warren, NJ) endovascular stent was easily deployed, and the correct position was checked by direct contrast injection through the long sheath. This small group of patients has been compared to a group of 5 age-matched patients with coronary steal syndrome in whom the procedure has been performed with standard technique including femoral approach and guide catheter. Results The procedure was successful in all patients; vertebral and LIMA ostia remained patent in all cases. In the control group, cannulation of the subclavian artery was difficult in two cases, while one patient developed a groin hematoma. Mean pretreatment gradient was 32 mm Hg with a range of 25 to 40 mm Hg (34 mmHg, range 26-43, in the control group, P=0.87) and fell to 2 mm Hg with a range of 0 to 4 mm Hg (3.1 mmHg, range 0 to 5, P=0.89) posttreatment. Mean contrast dose was 60±16 ml (138±26 ml in the control group, P＞0.01), whereas mean fluoroscopy and procedural time were 5.7±1.6 minutes (10.8±1.0 minutes in the control group, P＞0.01) and 15.7±6.3 minutes (28±7.1 minutes in the control group, P＞0.01). At a mean follow-up of 10±3.2 months all patients are alive and free from angina and residual induced ischemia. Conclusions Our brief study suggested that brachial artery access be considered the optimal route to treat coronary-subclavian steal syndrome in elderly patients because of clear advantages; these included no manipulation of catheter to cannulate the artery, perfect coaxial position of the catheter at the site of LSA stenosis, clear visualization of the LIMA and vertebral ostia, and easy access to these vessels in case of plaque shifting or embolic protection device deployment.