OBJECTIVE. The purpose of this study was to determine whether a visually apparent increase in the anteroposterior diameter of the spinal canal ('wide canal sign') can be used reliably to differentiate degenerative from isthmic spondylolisthesis on midline sagittal MR images. We hypothesized that the wide canal sign would be present only in isthmic spondylolisthesis, where the vertebral ring is disrupted by defects in the pars interarticularis. MATERIALS AND METHODS. The midline sagittal MR images of 100 control subjects without spondylolysis or spondylolisthesis were analyzed to establish the normal range of sagittal canal diameters at the various lumbar levels. Midline sagittal MR images of the lumbar spine were reviewed in 53 patients in whom either isthmic (n = 35) or degenerative (n = 18) spondylolisthesis was confirmed with conventional radiography and/or CT. For each spinal level, the sagittal canal ratio, defined as the maximum anteroposterior diameter of the canal at that level divided by the diameter of the canal at L1, was calculated. From our analysis of the control subjects, a normal sagittal canal ratio was determined to be less than 1.25. A sagittal canal ratio of 1.25 or greater at the level of a spondylolisthesis was considered to represent an abnormally increased sagittal canal diameter (wide canal sign), indicating the presence of bilateral pars interarticularis defects. Using this sign alone, two neuroradiologists who had no knowledge of the true diagnosis classified the type of spondylolisthesis in a randomized subgroup of 34 age-matched adults (all more than 40 years old and with grade I isthmic or degenerative spondylolisthesis). RESULTS. In the 100 subjects without spondylolisthesis and in the 18 patients with degenerative spondylolisthesis, the sagittal canal ratio did not exceed 1.25 (mean values, 0.93-0.99) at any level of the lumbar spine. Conversely, the sagittal canal ratio at the level of isthmic spondylolisthesis exceeded 1.25 (mean value, 1.56) in 34 of 35 patients. In patients more than 40 years old, in whom degenerative spondylolisthesis is prevalent and misdiagnosis of isthmic spondylolisthesis is potentially more likely, both neuroradiologists were 100% accurate in correctly assigning the type of spondylolisthesis when using the wide canal sign alone. CONCLUSION. The wide canal sign on midline sagittal MR images (corresponding to an sagittal canal ratio ≥ 1.25) is a reliable predictor of the presence of defects of the pars interarticularis at the level of a spondylolisthesis. This sign could be useful for distinguishing degenerative from isthmic spondylolisthesis when degenerative sclerosis in the pars interarticularis mimics spondylolysis, when direct axial imaging of the pars interarticularis has not been done, or when imaging is technically inadequate.