Background: Bleeding events immediately after PCI are influenced by multiple therapies and risk factors. The relation of platelet reactivity (PR) to post-discharge bleeding events in PCI-treated patients is unclear. Methods: Stable patients (n=257) without in-hospital bleeding were prospectively enrolled. Platelet measures were assessed in-hospital and at 30-day follow-up by conventional aggregometry and VASP-P assay. Primary end point was the correlation of 30-day incidence of bleeding events assessed by BleedScore��� (Superficial, Internal and Alarming) with PR. Results: A total of 73 patients experienced bleeding events (28.4%), of which 21.4% were superficial, and 7.0% were internal. Baseline demographics and in-hospital PR did not significantly differ across the bleeding type. Compared to patients without bleeding and experiencing superficial bleeding, those with an internal bleeding episode showed significantly lower PRs at 30-day (VASP-P index, 57.8짹17.6% vs. 45.7짹19.5%, p=0.007; 20µM ADP-PR, 50.5짹18.4%vs. 35.9짹23.3%, p=0.002; 6µg/ml collagen-PR, 45.8짹19.4% vs. 35.6 짹26.7%, p=0.037; 0.5mM arachidonic acid-PR, 17.4짹11.1% vs.14.7짹9.6%, p=0.332). By ROC curve analysis, the optimal cutoffs for predicting internal bleeding were 20µM ADP-PR at 30-day ���44.5% (AUC 0.663, 95% CI 0.525 to 0.802, p=0.016) and VASP-P index at 30-day ���43.1% (AUC 0.640, 95% CI 0.506 to 0.774, p=0.043). In multivariate analysis, 30-day cutoff of VASP-P (OR 7.0, 95% CI 1.8 to 27.6, p=0.005) or 20µM ADP-PR (OR 3.6, 95% CI 1.0 to 12.4, p=0.047) was an independent determinant of internal bleeding. Conclusion: Among patients undergoing PCI, on-treatment PR at 30-day predicts the risk of post-discharge internal bleeding, of which clinical impact needs to be validated in the future trials.