Objective: The aim was to intraoperatively identify the spinal cord blood supply and shorten the aortic crossclamp time, Methods: A platinum electrode was placed intrathecally by lumbar puncture alongside the spinal cord, After the aorta was crossclamped, hydrogen in a saline solution was injected into the aorta and, if it was shown that the segment supplied the spinal cord and there were multiple arteries, then these were individually injected. The repair was performed by a sequential segmental method as described previously, Results: Postoperatively, highly selective angiography was used to confirm that reattached intercostal arteries supplied the spinal cord, The technique was accurrate in all patients. Five spinal cord perfusion patterns were noted: (1) direct, (2) collateral, (3) no direct supply from segment tested, (4) from atriofemoral bypass, and (5) occluded reattached intercostals, When no response was obtained or no further testing was required (n = 8), testing time was 4.2 minutes and crossclamp time 41.9 minutes, When multiple segmental arteries required further testing, the mean testing time was 10.4 minutes and crossclamp time 58.5 minutes, including reattachment of intercostal vessels (p = not significant), Conclusion: Preliminary findings indicate that this method is a safe research technique, can detect radicular arteries, and may reduce the time for aortic crossclamping if no vessels are identified as supplying the spinal cord.