BACKGROUND: Antithrombotic drug users are at risk of intracerebral hemorrhage (ICH). Cerebral microbleeding (MB) is also a risk factor for ICH. Therefore, it is important to clarify the relationship between antithrombotic therapy (AT) and MB in order to prevent ICH. The incidence and number of MBs were evaluated in AT users and nonusers among consecutive acute ICH cases. We attempted to gain insights into the statistical“ black box” of hemorrhagic phenomena in which AT, MB, and ICH are mutually related. METHODS: T2*-weighted magnetic resonance imaging (1.5T MRI, 25 brain slices, 2-mm thick) was performed in 273 of 384 consecutive acute ICH (mean age 69.4yrs; M:F, 166:107) to evaluate the incidence and number of MBs, excluding the low-signal intensities associated with the hematoma lesions. RESULTS: Of the 273 cases , 86 (32%) cases were users and 187(68%) were nonusers, and MBs were detected in 60 (70%) users and 116 (62%) non-users. The incidence of MBs was comparable between users and nonusers (Chi-square independence test: χ2, 2.018 ; χ2(0.95)3.842; P = 0.155; odds ratio, 0.67129). The mean number of MBs was also not significantly different between users and nonusers [13.03 (median, 7) vs. 12.46 (median 7): t test: t -0.208; t (0.95) 1.653: P = 0.582]. CONCLUSION AT did not increase MB in this study patients, suggesting that this treatment does not contribute to vascular wall fragility. Complex phenomena leading to ICH were simply categorized into 3 steps: fragility of vascular wall, breakdown of vascular wall, and extension of hemorrhage. Antithrombotic therapy affects only the 3rd step and do not affect the 1st and the 2nd step, whereas hypertension affects all 3 steps. Therefore, in order to prevent hemorrhage extension and severe brain damage, stricter blood pressure control is required in antithrombotic drug users with MBs than in nonusers with MBs.