Objectives/Hypothesis: Systematically review the outcomes of endoscopic endonasal techniques to reconstruct large skull base defects (ESBR). Such surgical innovation is likely to be reported in case series, retrospective cohorts, or case-control studies rather than higher level evidence. Study Design: Systematic review and meta-analysis. Methods: Embase (1980-December 7, 2010) and MEDLINE (1950-November 14, 2010) were searched using a search strategy designed to include any publication on endoscopic endonasal reconstruction of the skull base. A title search selected those articles relevant to the clinical or basic science of an endoscopic approach. A subsequent abstract search selected articles of any defect other than simple cerebrospinal fluid (CSF) fistula, sella only, meningoceles, or simple case reports. The articles selected were subject to full-text review to extract data on perioperative outcomes for ESBR. Surgical technique was used for subgroup analysis. Results: There were 4,770 articles selected initially, and full-text analysis produced 38 studies with extractable data regarding ESBR. Of these articles, 12 described a vascularized reconstruction, 17 described free graft, and nine were mixed reconstructions. Three had mixed data in clearly defined patient groups that could be used for meta-analysis. The overall CSF leak rate was 11.5% (70/609). This was represented as a 15.6% leak rate (51/326) for free grafts and a 6.7% leak rate (19/283) for the vascularized reconstructions (Ï‡ 2 = 11.88, P =.001). Conclusions: Current evidence suggests that ESBR with vascularized tissue is associated with a lower rate of CSF leaks compared to free tissue graft and is similar to reported closure rates in open surgical repair.