BACKGROUND: The purpose of this study is to identify the differences of risk factors and stroke mechanism between early and late recurrence in patients with long-term antiplatelet therapy for stroke prevention. METHODS: We enrolled 114 consecutive patients with recurrent infarction who had been taking antiplatelet agents regularly since previous noncardioembolic cerebral infarction. Total 81 patients (49 men and 32 women) were met to the inclusion criteria through standardized evaluation. Subjects were classified into two groups depending on the time-to-recurrence after antiplatelet therapy: early antiplatelet failure (within 2 years, n=41, hereafter as "EAF") and later antiplatelet failure (after 2 years, n=40, hereafter as "LAF"). We investigated the differences of clinical factors between two groups using univariate and multivariate analysis. RESULTS: Family history of stroke (29.3% in EAF vs. 10% in LAF, p=0.029) was more frequent in EAF group. Low HDL-cholesterol and High total cholesterol/HDL-cholesterol ratio were associated with the LAF group (p=0.042, 0.005 respectively). Multivariate analysis showed that family history of stroke (OR=5.283, 95%CI 1.178-23.699, p=0.030) and previous infarction classified as large artery atherosclerosis (OR=8.497, 95%CI 1.444-50.015, p=0.018) were significant predictors for EAF whereas total cholesterol/HDL-cholesterol ratio (OR=2.002, 95%CI 1.183-3.389, p=0.010) was for LAF. CONCLUSIONS: This study suggests that family history of stroke and cerebral infarction due to large artery atherosclerosis are more responsible for the early recurrence while dyslipidemic condition is more related to the late recurrence during long-term antiplatelet therapy in patients with previous cerebral infarction.