Aim To explore the association between insomnia and myocardial infarction (MI) using bidirectional two-sample Mendelian randomization (MR). Methods The summary statistics of MI were obtained from genome-wide association studies (GWAS), and single nucleotide polymorphisms (SNP) associated with insomnia phenotype were selected as instrumental variables. Inverse variance weighting (IVW) was used to evaluate the causal relationship between insomnia and MI. Results MR analysis showed that there was positive relationship between insomnia and MI (IVW:OR=1.5,5%CI 1.004～1.027, P＝0.007). There was no horizontal pleiotropy and heterogeneity in instrumental variables. Sensitivity analysis suggested that MR analysis results were stable. Reverse MR analysis showed no causal relationship between MI and increased risk of insomnia (IVW:OR=0.5,5%CI 0.466～1.126, P＝0.152), and the statistical results were stable after the removal of outliers. Conclusion Insomnia is a risk factor for MI and there was positive relationship between insomnia and MI. However, there was no causal relationship between MI and an increased risk of insomnia.