Background: Aortic intramural hematoma (AIH), a variant form of classic aortic dissection (AD), has been accepted as an increasingly recognized and potentially fatal entity of acute aortic syndrome (AAS). The incidence of AIH is not known, and predictors of death for whole patients with type A AAS were not seriously investigated. Methods and Results: We evaluated 186 consecutive patients with type A AAS (57.0짹13.5 years, 95 females) admitted from January 1993 to March 2003. Final diagnosis was AD in 135 patients and AIH in 51(27%). Patients with AD were younger (54.0짹13 vs 65.6짹10.7, p<0.05) and female gender was more prevalent in AIH group (p<0.05). Overall in-hospital mortality was 16.1% (30/186); Mortality of AD group was higher than that in AIH (26/135 vs 4/51, p=0.059), and mortality without surgery was also significantly higher in AD group (17/24 vs 3/35, p<0.001). Logistic regression identified the following presenting variables as predictors of mortality: AD (OR 22.92; 95% CI, 3.14 to 167.16; p<0.005), confusion (OR 17.90; 95% CI, 3.90 to 82.06; p<0.001), shock/tamponade (OR 9.39; 95% CI, 2.58 to 34.06; p<0.001), and dyspnea/heart failure (OR 10.77; 95% CI,1.77 to 65.64). Shock/tamponade was more prevalent in AIH group than in AD(12/51 vs 15/135, p=0.032), and shock/tamponade was a predictor of higher mortality in both groups. Pericardiocentesis for shock/tamponade was more frequently done in AIH group than in AD (7/12 vs 3/15, p<0.05), and mortality was lower in AIH group (3/12 vs 10/15, p<0.05). Conclusions: AIH comprises of significant proportion of type A AAS and is an independent variable associated with lower mortality. Shock/tamponade is a challenging presenting variable in both AD and AIH, and role of pericardiocentesis in AIH should be reevaluated.