Spinal cord ischemia (SCI) remains a significant concern in patients undergoing endovascular repair involving the thoracic aorta (thoracic endovascular aortic repair [TEVAR]). Perioperative lumbar spinal drainage has been widely practiced for open repair, but there is no consensus treatment protocol using lumbar drainage for SCI associated with TEVAR. This study analyzes the efficacy of an institutional protocol using selective lumbar drainage reserved for patients experiencing SCI following TEVAR. A prospectively maintained registry was reviewed to identify all patients who underwent TEVAR from January 2000 through June 2010. Preoperative characteristics, intraoperative details, and outcomes, including neurologic deficit and mortality at 30 days and 1 year were determined based on reporting standards. Patients developing symptoms of SCI in the postoperative setting were compared with those without neurologic symptoms. SCI patients who received selective lumbar drainage were grouped based on resolution of neurologic function, with risk factors and outcomes of these subgroups analyzed with χ 2, t test, logistic regression, and analysis of variance (ANOVA). Two hundred seventy-eight TEVARs were performed on 251 patients. Twelve patients accounting for 12 TEVARs were excluded from analysis: 5 patients experienced SCI preoperatively, 4 patients were drained preoperatively, 2 expired intraoperatively, and 1 procedure was aborted. Of the remaining 266 procedures in 239 patients, 16 (6.0%) developed SCI within the 30-day postoperative period. Risk factors for SCI reaching statistical significance included length of aortic coverage (P =.036), existence of infrarenal aortic pathology (P =.026), and history of stroke (P =.043). Stent graft coverage of the left subclavian artery origin was required in 28.9% (n = 77) and was not associated with SCI (P =.52). Ten of 16 post-TEVAR SCI patients received selective postoperative lumbar drains and were categorized based on resolution of symptoms into complete resolution (n = 3; 30%), partial resolution (n = 4; 40%), and no resolution (n = 3; 30%). No patient characteristics or risk factors reached significance in comparison of lumbar drained patients and nondrained patients. All seven drained patients without complete resolution of SCI died within the first year after surgery, while all three of the complete responders survived (P =.017). In patients with SCI, increased all-cause mortality was observed at 1 year (56.3% vs 20.4%; P =.003). A protocol utilizing selective postoperative lumbar spinal drainage can be used safely for patients developing SCI after TEVAR with acceptably low permanent neurologic deficit, although overall survival of patients experiencing SCI after TEVAR is diminished relative to non-SCI patients. © 2012 Society for Vascular Surgery.