Background The previous large‐scale randomized controlled trial showed that routine thrombus aspiration (TA) during percutaneous coronary intervention (PCI) was associated with an increased risk of stroke. However, real‐world clinical evidence is still limited. Methods and Results We investigated the association between manual TA and stroke risk during primary PCI in the OACIS (Osaka Acute Coronary Insufficiency Study) database (N=12 093). The OACIS is a prospective, multicenter registry of myocardial infarction. The primary end point of the present study is stroke at 7 days. A total of 9147 patients who underwent primary PCI within 24 hours of hospitalization were finally analyzed (TA group, n=4448, versus non‐TA group, n=4699 patients). TA was independently associated with risk of stroke at 7 days (odds ratio [OR], 1.92 [95% CI, 1.19‒3.12]; P =0.008) in the simple logistic regression model, while the multilevel random effects logistic regression model with hospital treated as a random effect showed that manual TA was not associated with incremental risk of stroke at 7 days (OR, 0.91 [95% CI, 0.71‒1.16]; P =0.435). The 7‐day stroke risk of manual TA was significantly heterogeneous in different institutions ( P _{for interaction} =0.007). Conclusions Manual TA during primary PCI for patients with acute myocardial infarction was independently associated with the overall increased risk of periprocedural stroke. However, this result was substantially skewed because of institution specific risk variation, suggesting that the periprocedural stroke may be preventable by prudent PCI procedure or appropriate periprocedural management. Registration URL: https://upload.umin.ac.jp/cgi‐open‐bin/ctr_e/ctr_view.cgi?recptno=R000005464 . Unique identifier: UMIN000004575.