BACKGROUND This study investigated whether the prolongation of the cauda equina conduction time (CECT) was related to the type of neurogenic intermittent claudication due to lumbar spinal stenosis. METHODS In total, 149 patients who underwent surgery due to lumbar spinal stenosis with neurogenic intermittent claudication were classified into three groups as follows: cauda equina-type(n = 67), radicular-type(n = 29), and mixed-type(n = 53). Cauda equina conduction time was measured by placing disc electrodes on the abductor hallucis muscle, electrically stimulating the tibial nerve of the ankle and recording the compound muscle action potentials and F-waves. Motor evoked potentials from the abductor hallucis muscle were measured after magnetically stimulating the lumbosacral spine. Cauda equina conduction time was calculated from the latencies of compound muscle action potentials, F-waves, and motor evoked potentials. The measurement of the dural sac cross-sectional area were assessed using computed tomography myelography or MRI. RESULTS The values of CECT were as follows: cauda equina-type, 5.6 ± 1.1 ms; mixed-type, 5.1 ± 0.9 ms; and radicular-type, 4.0 ± 0.9 ms. The values of dural sac cross-sectional area were as follows: cauda equina-type, 42.8 ± 18.7 mm; mixed-type, 49.6 ± 20.9 mm; and radicular-type, 75.3 ± 19.1 mm. In the cauda equina-type and mixed-type patients, CECT was significantly prolonged and there were negative correlations between CECT and dural sac cross-sectional area. CONCLUSIONS Cauda equina conduction time differed according to the type of lumbar spinal stenosis. The prolongation of CECT may be caused by the demyelination of the CE. Cauda equina conduction time may be a useful measure for evaluating the dysfunction of the CE rather than radiculopathy for patients with lumbar spinal stenosis.