Spinal canal dimensions are assumed to play a significant role with regard to the development of symptoms in individuals with disc herniations. The literature is inconclusive on the significance of spinal canal size as a risk factor for sciatica, mainly because of study design problems. The objective of this study, therefore, was to test the hypothesis that spinal canal dimensions are a significant risk factor for the development of sciatica, comparing symptomatic and asymptomatic individuals. Thirty symptomatic patients undergoing lumbar discectomy and 45 asymptomatic volunteers were investigated by clinical and MRI examination. The size of the spinal canal and thecal sac as well as the midsagittal spinal canal diameter were measured using a point counting method and scanner software, respectively. Differences between the groups were compared separately for each level L3/4 to L5/S1. The intra- and inter-observer error ranged between 0.95 and 0.99 for all measurements. In symptomatic patients, the dimensions of the spinal canal and thecal sac as well as the midsagittal spinal canal diameter were smaller at all disc levels. Unpaired t-test demonstrated a significant difference, ranging from P<0.05 to P<0.001. When controlled for age, sex and body height, the odds ratio for a symptomatic disc herniation increased to as high as 35, depending on the spinal level, when the size of the spinal canal was smaller than the mean for controls by two standard deviations or more. In symptomatic patients, spinal canal dimensions are significantly smaller than those in asymptomatic individuals. Spinal canal dimension is an important factor discriminating patients from control subjects. A clinically relevant grading system for disc herniation should therefore be based on the spatial relationship between herniated disc material and neurogenic structures.