Lumbar spine foraminal stenosis has previously been defined by the foraminal and posterior disc height. We performed a study to determine whether residual clinically significant foraminal stenosis correlates with foraminal dimensions and peri-neural fat signal loss in a group of patients with leg pain undergoing surgery for lateral recess stenosis. We retrospectively studied the pre-operative para-sagittal MRI slices of 57 patients undergoing lumbar decompression and measured pre- and post-operative VAS as a primary outcome measure to evaluate surgical success. We performed a linear regression analysis comparing change in VAS score, 1 year VAS and percentage change in VAS with foraminal height and width and found no significant correlation (R² <0.2 for all correlations). We identified a sub-group of 16 patients with absent perineural foraminal fat signal with a significantly increased probability of post-operative VAS>2 compared to patients with present fat signal (p=0.0001) who all had foraminal height <10mm. In conclusion, we were unable to define dimensional foraminal parameters for clinically significant foraminal stenosis on para-sagittal MRI. Obliteration of perineural fat was associated with worse outcome and post-operative leg pain. The aetiology of foraminal stenosis is multi-factorial and more detailed imaging of the foramen is required. We recommend that coronal and fine para-sagittal MRI slices are analysed to evaluate patients with central and lateral recess stenosis for co-existing foraminal stenosis.