Background: Periventricular hyperintensity (PVH), deep and subcortical white matter hyperintensity (DSWMH), and silent brain infarct (SBI) have been proposed as subclinical risk markers for future symptomatic stroke onset. However, the differences between PVH, DSWMH, and SBI as markers for eventual development of symptomatic stroke in middle-aged healthy Japanese individuals has not been adequately explored. Methods: We examined 280 Japanese adults (92 female, mean age, 52.9±5 years) with no history of cardio-cerebrovascular disease. PVH, DSWMH, SBI and various vascular risk factors for each subject were assessed, and 10-year stroke risk was predicted using the point-based prediction model provided by the Japan Public Health Center (JPHC) study, which consists of six risk factors that are used to calculated predicted stroke risk, including age, sex, smoking status, body mass index (BMI), blood pressure (BP), and diabetes status. Results: The subjects were divided into four groups based on predicted stroke risk: low, <1%; low-medium, 1- <5%; medium-high, 5-<10%; high, ≥10%. According to the JPHC study risk-prediction algorithm, a PVH grading ≥ 2, a DSWMH grading ≥2, and a SBI number ≥1 were positively correlated with an increase in predicted stroke risk. Age, systolic BP, HbA1c levels, eGFR, hypertension, diabetes mellitus, and family history of cardio-cerebrovascular disease (CCVD) were found to be significantly associated with SBI. The odds ratio (OR) for high predicted stroke risk (10-year risk ≥10%) showed significance for SBI ≥1, even after adjusting for potential confounding factors (OR, 12.45: 95% CI, 3.09-50.84), whereas the ORs for PVH and DSWMH were not significant. Conclusion: The presence of an SBI, not white matter lesions, was significantly correlated with an increased 10-year stroke risk, suggesting that detection of SBI by MRI may be practical for identifying a high risk of stroke for middle-aged Japanese subjects.