Thirty-two patients with suspected lumbar disc herniation were studied with magnetic resonance imaging (MRI) and contrast computed tomography (CT). One hundred disc levels were evaluated. Twenty-five patients underwent surgery on 31 discs, allowing anatomic confirmation of the diagnosis. Surgical findings supported the MRI diagnosis at 28 of 31 levels (90.3% accuracy), whereas the CT diagnosis correctly reflected only 24 of 31 levels (77.4% accuracy). Discrepancy between MRI and CT interpretation occurred at ten levels that were surgically explored. Computed tomography (CT) was incorrect at seven levels, and MRI was in error at three levels. The sensitivity of MRI was 91.7%, compared with 83.3% for CT, and the MRI specificity of 100% was superior to 71.4% for CT. This study demonstrates the clinical superiority of surface coil MRI over contrast CT in the evaluation of lumbar disc herniation. Surface coil MRI can be used as the initial diagnostic procedure for a suspected herniated lumbar disc, using invasive contrast studies and CT, if required, to clarify an equivocal MRI finding.