•1-year retrospective chart review was conducted for thoracic epidurals attempted at a single quaternary care academic medical center.

•Available CT images were systematically reviewed to evaluate anatomical factors associated with technically difficult placement

•This analysis was performed using standard 2D CT scans (three planes) and 3D-rendered images.

•A suggested best approach for epidural placement was also recorded, based on the most open space on imaging.

•There were 11 epidural placement failures among 524 total thoracic epidurals attempted (98% success rate).

•For patient characteristics, the majority were male, with varying ages and morphometrics.

•Upper abdominal surgery was the most common indication.

•The most common anatomic findings: apposition of spinous processes (< 2 mm) (n = 8, 73%); scoliosis/rotation of the spine (n = 6, 55%); and narrow interlaminar spaces (< 3 mm) (n = 5, 45%) (Table 2).

•The 3D CT renderings facilitated identification of relatively open versus closed spaces (Images 1 and 2) in each case.

•Midline trajectory analysis frequently predicted more bony contact compared to paramedian

•In 10 (91%) of the failed epidurals, CT analysis revealed at least one potentially viable approach at an appropriate level

•This study demonstrates the feasibility of spine analysis to identify optimal interspaces and anatomical contributors to epidural difficulty, with challenging or failed cases often showing indicators detectable through combined 2D and 3D evaluation.

•3D trajectory analysis may identify interspaces where epidural placement is likely futile, supporting earlier selection of alternate levels or alternative regional analgesic techniques.