The aim of this study was to explore whether matched or mismatched strains of influenza vaccines (IVs) are beneficial at reducing the risk of acute myocardial infarction (AMI) in elderly persons.Data were obtained from the Longitudinal Health Database 2005 (LHID 2005) which is maintained by the National Health Insurance Research Institute in Taiwan. The analytical data included individuals who were vaccinated with mismatched vaccines during the October 2007 to December 2007 season and individuals vaccinated with matched strains during the October 2008 to December 2008 season. All participants were 65 years of age and older. In this analysis, individuals were considered to be exposed if their records showed that they were vaccinated against influenza, and they were considered to be nonexposed if they were not vaccinated during these seasons. A Cox hazard model was used to estimate AMI hazard ratio.This study enrolled 93,051 exposed and 109,007 unexposed individuals. The AMI hazards ratios (HRs) for the men and women exposed to mismatched vaccine (in 2007) were 0.990 (95% confidence interval [CI], 0.745-1.316) and 1.102 (95% CI: 0.803-1.513), respectively. Men exposed to matched vaccines (in 2008) had significant HRs (HR: 0.681; 95% CI: 0.509-0.912) while the HRs in the women were barely significant (HR: 0.737; 95% CI: 0.527-1.029).AMI risk could be particularly reduced in men if the IV matches well with the circulating strains in elderly people 65 years of age and older.