BACKGROUND: Low back pain (LBP) and radicular leg pain (RLP) are among the most common types of pain in human beings. Although magnetic resonance imaging (MRI) is very sensitive for diagnosis of discopathy, some factors, such as overestimation of pathology, expensiveness, unavailability, and using it for patients with cardiac pacemaker or metal foreign bodies, limit the utility. The present study is designed to evaluate the efficacy of computed tomography scan (CTS) in patients with disc herniation in each level of Lumbar spine versus MRI findings at the same level.METHODS: In a prospective trial, 100 consecutive patients with LBP and RLP and signs and symptoms of discopathy referred to our private clinic from September 2004 to April 2005 were studied. CTS and MRI and their data were compared level by level; i.e. CTS of the patients analyzed according to clinical signs and symptoms and compared with MRI at the same level in axial view.RESULTS: Thirty-two patients had clinically S1 root signs and symptoms, in all of them CTS and MRI showed disc herniation at L5/S1 level in axial view. For L5/S1 level, positive predictive value (PPV) of CTS was 100%. In upper Lumbar region, CTS findings were less reliable than MRI. CTS showed the pathology at 14.2% of upper Lumbar, 27.2% at L3/L4 and 46.3% at L4/L5. In nine cases with more than one level involved, CTS confirmed the diagnosis in 11.1% of the cases.CONCLUSIONS: MRI is the gold standard for diagnosis of Lumbar disc herniation, but CTS is sensitive in 100% for L5/S1, 68% for L4/L5, 60% for L3/L4, 0% for upper Lumbar discopathies and finally 78% for multilevel involvement. Therefore, the higher the level of disc herniation is, the lower the sensitivity of CTS.