Background Several studies have suggested that periprocedural myocardial infarction (MI) following percutaneous coronary intervention (PCI) is associated with worse prognosis, but a definite study is not available. We therefore evaluated the incidence, mechanism, and the long-term prognostic impact of periprocedural MI in a large number of PCI patients. Methods We analyzed individual patient-level data of studies (randomized clinical trials or registries) of patients undergoing PCI with stent implantation in which postprocedural creatine kinase-MB (CK-MB) and mortality data were collected. Periprocedural MI was defined as an elevation of the CK-MB isoenzyme at least 3 times the upper limit of normal in at least 2 blood samples within 24 hours after PCI. Results A total of 23,002 patients from 10 studies were included. Overall incidence of periprocedural MI was 6.7%. Follow-up varied from 1 year to 5 years. During follow-up, the presence of periprocedural MI was significantly associated with long-term mortality (hazard ration [HR] 1.49, 95% confidence interval [CI] 1.20-1.86, P<0.001). Mortality was found to be a monotonically increasing function of the CK-MB ratio; ratio 3-5 with HR of 1.33, ratio 5-10 with HR of 1.63, and ratio >10 with HR of 1.84. However, a periprocedural MI was not prognostically equivalent to a spontaneous MI for long-term mortality (spontaneous MI vs. periprocedural MI; HR 3.61, 95% CI 2.64-4.94, P<0.001). The key predictors of periprocedural MI were age, gender, diabetes, acute coronary syndrome, multivessel disease, bifurcation, and stent length. Most common mechanism of periprocedural MI was a side-branch occlusion. Conclusion Among patients who had undergone PCI with stent implantation, the frequency of periprocedural MI was not rare, and mainly related to side-branch occlusion. The presence of periprocedural MI was associated with long-term risk of mortality, but the degree of association was less than spontaneous MI.