A prospective study using a newly developed imaging diagnosis method to examine the patients with L5 radiculopathy. To confirm the efficacy of a new 3-dimensional computed tomography (3D CT) imaging method to diagnose extraforaminal stenosis at the lumbosacral junction. The lack of recognition of extraforaminal stenosis at the lumbosacral junction is a cause of failed back surgery syndrome. A clear method to diagnose this condition is essential. The participants were 75 consecutive patients (mean age 69.5 y) with or without cauda equina symptoms who were treated by microendoscopic spinal surgery for L5 radiculopathy. The lesion responsible for the symptoms was identified by a combination of neurologic findings, selective radiculography, (3D MRI), and intraoperative neurophysiological findings. Multislice CT scanning was carried out preoperatively from the L1 vertebral body to the sacrum in all patients. The CT scan images were transferred to a remote computer workstation, and the reconstructed images were examined after surgery by an investigator blinded to the clinical diagnoses. In the reconstructed plane, we measured the minimum cross-sectional area of the de novo bony tunnel formed by the L5 transverse process, sacral ala, and L5 vertebral body, that is, lumbosacral bony tunnel (LSBT) and determined a cutoff value to diagnose extraforaminal stenosis. The shape of the LSBT was also evaluated in relation to the diagnosis. In 3D CT analysis, the LSBT was found on the ipsilateral side in 51 of the 75 patients. The bony tunnel was outside the foramen in all patients diagnosed clinically with extraforaminal stenosis, but in only 60% of the patients without extraforaminal stenosis. The minimum cross-sectional area of the bony tunnel was significantly smaller in patients with an extraforaminal stenosis than in those without extraforaminal stenosis. The cutoff value was set at 0.8 cm2. The cross-sectional area was <0.8 cm2 in all patients with extraforaminal stenosis. The specificity of this diagnostic procedure was 89.6%, and the sensitivity was 100%. All true-positive cases had the spur-type shape of the bony tunnel, and all false-positive cases had the round-type shape. All patients with extraforaminal stenosis had an LSBT. The minimum cross-sectional area of the bony tunnel was significantly smaller in patients with an extraforaminal lesion than in those without an extraforaminal lesion. 3D CT is a useful tool for diagnosing extraforaminal stenosis at the lumbosacral junction.