Lipoprotein(a) [Lp(a)], which is genetically determined, has been reported as an independent risk factor for atherosclerotic vascular disease. However, the prognostic value of Lp(a) for secondary vascular events in patients after coronary artery disease has not been fully elucidated. This 3-year observational study included a total of 176 patients with ST-elevated myocardial infarction (STEMI), whose Lp(a) levels were measured within 24 h after primary percutaneous coronary intervention. We divided enrolled patients into two groups according to Lp(a) level and investigated the association between Lp(a) and the incidence of major adverse cardiac and cerebrovascular events (MACCE). A Kaplan–Meier analysis demonstrated that patients with higher Lp(a) levels had a higher incidence of MACCE than those with lower Lp(a) levels (log-rank P = 0.034). A multivariate Cox regression analysis revealed that Lp(a) levels were independently correlated with the occurrence of MACCE after adjusting for other classical risk factors of atherosclerotic vascular diseases (hazard ratio 1.030, 95 % confidence interval: 1.011–1.048, P = 0.002). In receiver-operating curve analysis, the cutoff value to maximize the predictive power of Lp(a) was 19.0 mg/dl (area under the curve = 0.674, sensitivity 69.2 %, specificity 62.0 %). Evaluation of Lp(a) in addition to the established coronary risk factors improved their predictive value for the occurrence of MACCE. In conclusion, Lp(a) levels at admission independently predict secondary vascular events in patients with STEMI. Lp(a) might provide useful information for the development of secondary prevention strategies in patients with myocardial infarction.