Purpose: To evaluate the incidence and natural history of endoleaks following thoracic endovascular aortic repair (TEVAR) of thoracic aortic aneurysm (TAA). Methods: A retrospective review was conducted on 200 TAA patients (122 men; mean age 69±12 years) who underwent elective TEVAR in the descending thoracic aorta between January 2001 and December 2006. The mean aneurysm diameter was 66 mm (range 50-124), and most (75%) of the aneurysms were atherosclerotic in origin. Medical records and multidetector computed tomography studies were evaluated on a 3-dimensional workstation to categorize endoleak, establish morphological characteristics, and assess outcomes. The outcomes following any secondary interventions were noted in the context of endoleak etiology and the magnitude of the intervention. Results: Over a mean follow-up of 30 months (range 12-73), 39 (19.5%) patients developed an endoleak (33 primary and 8 secondary); 3 people had 2 distinct types of endoleaks. Endoleaks were associated with the presence of a carotid-subclavian bypass (p=0.0001) and lengthy aortic coverage by the stent-graft (p=0.005). The proportion of the 170 patients with a Zenith stent-graft who had an endoleak (17%, n=39) was significantly lower (p<0.01) than the proportion of endoleaks (34%, n=10) in the 30 patients with Gore or Talent devices. Secondary interventions for endoleak were performed in 79% of type I, 24% of type II, and 57% of type III endoleaks. Conclusion: In this study, 1 in 5 TAA patients with TEVAR had endoleak. Most type I and III endoleaks required secondary intervention, while conservative treatment was most frequent for type II. Characterization of endoleak type was not always precise, and routine surveillance of all patients with endoleak is recommended.