BACKGROUND Recent advances in best medical therapy (BMT) has been associated with reduced risk of stroke similar to that observed following surgical carotid revascularization (CR). Thus, it remains uncertain which subset(s) of patients would benefit from prophylactic CR+BMT for asymptomatic carotid stenosis (ACS) over BMT alone. The purpose of this study was to analyze the contemporary experience in the management of >70% ACS in an academic institution, to compare the short- and long-term outcomes of BMT alone against CR+BMT, and to identify risk factors for the development of future cerebrovascular events. METHODS A retrospective review of all patients with severe ACS between January 2005 and December 2012 at Loyola University Medical Center and its affiliated Edward Hines Jr. Veterans Administration Hospital was conducted. Baseline patient characteristics, medications, and follow-up data were collected from electronic medical records, and treatment outcomes were compared. The random forest method was performed to select potential important variables for the development of late stroke. The recursive partitioning regression analysis (RPRA) was performed to identify the patient subgroup at increased risk of future stroke. RESULTS Of 409 patients identified; 247 were treated with CR and 162 with BMT. Between these groups with CR+BMT and BMT alone, the mean age was 69.1±8.2 versus 75.5±9.0, respectively (P<0.01). Mean follow-up was 60.7±37.5 months. Early (30-day) outcomes of stroke, acute myocardial infarction or mortality did not differ between the treatment modalities (2.0% CR vs. 0.6% BMT, P=0.41). Probability of freedom from ipsilateral stroke, and any stroke at 1- and 5-year follow-up were also comparable between CR+BMT and BMT alone. However, random forest method and RPRA demonstrated that patients with history of diabetes and remote stroke treated with BMT alone were at a high risk for future stroke (36.4% in total, 7.2% per year). The diabetics with contralateral carotid stenosis >50% who are active smokers are at the highest risk for stroke after CR (20.0% in total, 4.0% per year). CONCLUSIONS Prophylactic CR+BMT does not provide overall late stroke prevention compared with BMT alone. Diabetics with a history of stroke, in particular, are at an increased risk of stroke despite BMT. Timely CR+BMT for high-risk patients is still indicated.