Background: There have been few reports about long-term prognosis of vasospastic angina (VA) in Korea. We estimated long-term mortality and rate of percutaneous coronary intervention (PCI) due to de novo atherosclerotic change in known VA using data-sets of VA-CMC (Vasospastic Angina in Catholic Medical Center) registry. Methods: Total 952 patients of 8 centers in CMC were enrolled to VA-CMC registry from March 2008 to October 2010. The patients were diagnosed as VA by acetylcholine or ergonovine provocation test using the same criteria. Patients who had significant atherosclerotic stenosis (>50% of luminal diameter narrowing) on their baseline angiography were excluded. We also excluded the patients with significant heart failure or renal failure, cancer, inflammatory disease, and catheter-induced spasm on angiography. All patients in the registry received conventional vasodilator therapy during follow up. Out of these patients, 442 patients were included as study subjects in this retrospective analysis. The mean follow-up duration was 39짹11 months. We evaluated the rate of cardiac or non-cardiac deaths and rate of PCI. Results: During follow up period, the rate of mortality was 2.71% (12/442). Nine patients died from cardiac cause (2.04%). Baseline and angiographic characteristics were similar between patients who died and survivors, however, patients who died demonstrated a higher frequency of diffuse type of vasospasm on initial spasm provocation test and a higher rate of drug cessation during follow up period than survivors (p<0.05, respectively). In multivariate analysis, drug cessation was an independent risk factor for cardiac mortality (odds ratio 1.47, p<0.05). The rate of PCI due to de novo atherosclerosis was 2.26% (10/442). Out of 10 patients, only 3 patients (0.68%) demonstrated the development of significant atherosclerosis on the segment of vasospasm documented on previous provocation test. In other 7 patients, PCI was done for de novo atherosclerosis on non-vasospastic segment. PCI group had higher numbers of male, current smoker, and higher level of initial hsCRP than non-PCI group (p<0.05, respectively). In multivariate analysis, current smoking (odds ratio 2.31, p<0.05) and high level of initial hsCRP (odds ratio 1.57, p<0.05) were independent risk factors for PCI. Conclusions: In Korean VA, the rates of cardiac mortality (2.04%) and development of de novo atherosclerosis (2.26%) were very low. During follow up, atherosclerotic change developed more frequently on non-vasospastic segment than on the segment of vasospasm. Current smoking, increased hsCRP and drug cessation were independent risk factors for unfavorable outcomes in Korean VA.