The Geriatric Nutritional Risk Index (GNRI) is a straightforward nutritional risk assessment tool with an established relationship with poor prognosis in patients with heart failure. However, the utility of the GNRI in patients with acute myocardial infarction (AMI) remains unclear given the time-dependent changes in the pathophysiology of AMI and the selected endpoints. Accordingly, we aimed to evaluate the optimal cut-off values of the GNRI for cardiovascular events in patients with AMI. We used time-dependent receiver operating characteristic analysis to identify the optimal cut-off values for two endpoints, all-cause death and major adverse cardiac events (MACE: all-cause death, non-fatal myocardial infarction, hospitalization for heart failure, and stroke), over 4 years in 360 patients with AMI between 2012 and 2020. The cumulative incidence of MACE was 11.6%. The cut-off value of the GNRI for all-cause death was 82.7 (area under the curve [AUC], 0.834) at 3 months and 90.3 (AUC 0.854) at 4 years. The cut-off value of the GNRI for MACE was 83.0 (AUC 0.841) at 3 months and 95.3 (AUC 0.821) at 4 years. The GNRI demonstrated consistently high reliability relative to other indicators of AMI. Our findings indicated that the optimal cut-off value and reliability of the GNRI for cardiovascular events varied according to the endpoints and observation periods. GNRI emerges as a crucial predictor of prognosis for patients with AMI.