Background and Objectives : Anesthesiologists are reluctant to consider higher levels for spinal anesthesia, largely due to direct threats to the spinal cord. The goal of this study is to investigate, with magnetic resonance imaging (MRI), the distances between the relevant structures of the spinal canal (spinal cord, thecal tissue, etc.) to determine modal anatomical positions for neuraxial anesthesia. Method : A group of 19 patients were imaged with an MRI scanner in supine position. Medial sagittal slices of the thoracic and lumbar spine were measured for the relative distances between anatomical structures, including epidural space, dura, and spinal cord. Results : The posterior dura – spinal cord distance is significantly greater in the middle thoracic region than at upper and lower thoracic levels (e.g. T6 9.5 ± 1.8 mm, T12 3.7 ± 1.2 mm, p < 0.001, T1 4.7 ± 1.7 mm, p < 0.001). There is variation in modal distances between the structures important for neuraxial anesthesia, at different levels of the spinal canal. Conclusions : The spinal cord tends to follow the straightest line through the imposed geometry of the spine. Considering the necessary angle of entry of the needle at mid-thoracic levels, there is relatively (more than at upper thoracic and lumbar levels) substantial separation of cord and surrounding thecal tissue. Anesthesiologists perform spinal blockades up to the L2-L3 interspace, but avoid higher levels for fear of neurological damage. The information that there is substantially more space in the dorsal subarachnoid space at thoracic level, might lead to potential applications in regional anesthesia. In contrast, the cauda equina sits more dorsally in the lumbar region.