BACKGROUND: Individuals with severe mental illnesses (SMI) like schizophrenia, bipolar disorder (BD), and major depressive disorder (MDD) have an increased risk for cardiovascular diseases (CVD), but the causal relationship remains unclear. METHODS: Mendelian randomization (MR) was used to investigate the potential causal relationship between SMI and CVD and its five subtypes of disease, coronary heart disease, myocardial infarction, stroke, heart failure, and atrial fibrillation. Subsequently, the MR results of SMI with CVD and its subtypes were meta-analyzed separately. To assess the robustness of the findings, Cochran's Q test, MR-Egger intercept test, MR-PRESSO, and leave-one-out analysis were used. Select single nucleotide polymorphisms (SNPs) related to SMI and CVD and their five subtypes (coronary heart disease, myocardial infarction, stroke, heart failure, and atrial fibrillation). Use univariable Mendelian randomization (UVMR) and multivariate Mendelian randomization (MVMR) to assess the causal relationship between these conditions. Conduct a meta-analysis of the MR results of SMI and CVD and their subtypes. Use MR mediation analysis to evaluate the mediating effect of BMI between BD and CVD. Use Cochran's Q test, MR-Egger intercept test, MR-PRESSO, and leave-one-out analysis to enhance the robustness of the study. RESULTS: MR analyses have revealed correlations between schizophrenia and BD with CVD and their subtypes in certain datasets. No significant evidence of an association between MDD and CVD or its subtypes was observed in our MR analyses. After MVMR and MR meta-analysis, no basis for genetically predicted SMI increasing CVD and their subtypes was found. The MR mediation analysis showed that the reduced risk of certain CVDs in BD was partially related to BMI to some extent. CONCLUSION: Our MR study did not provide conclusive evidence for a causal association between genetic predisposition to SMI and CVD. Based on the available evidence, it would be more appropriate to consider SMI as potential risk markers for CVD and its subtypes rather than definitive risk factors.