BACKGROUND AND AIMS: Studies have suggested that osteocalcin (OCN) is implicated in vascular calcification and linked to cardiovascular diseases (CVDs), but it is unclear whether the relationships are causal. The aim of this study is to evaluate the causal relationship of circulating OCN with CVDs and the role of vascular calcification. METHODS AND RESULTS: Bi-directional, mediation, and multivariable Mendelian randomization (MVMR) were performed using summary-level data for circulating OCN levels, coronary artery calcification (CAC), and CVDs, including coronary artery disease (CAD), myocardial infarction (MI), heart failure, atrial fibrillation, stroke and its subtypes. Pooled estimates from two independent datasets of OCN were calculated using the inverse variance weighted method with sensitivity analyses. The conservative Hochberg correction method adjusted the P-value for multiple comparisons. Genetically predicted higher OCN levels were linked to an increased risk of CAD (odds ratio [OR] = 1.069, 95%CI = 1.037-1.102, P <
0.001) and MI (OR = 1.099, 95%CI = 1.069-1.130, P <
0.001). In addition, elevated OCN levels were associated with higher CAC (β = 0.180, 95%CI = 0.101-0.258, P = 0.006), which was related higher risk of CAD (OR = 1.225, 95%CI = 1.132-1.325, P <
0.001) and MI (OR = 1.286, 95%CI = 1.203-1.375, P <
0.001), mediating 54.5 % and 48.3 % of the effect of OCN on CAD and MI, respectively. Meanwhile, MVMR results also validated the mediating role of CAC. In contrast, CAD and MI were associated with decreased levels of plasma OCN. CONCLUSION: Our findings reveal that higher OCN concentrations are associated with an elevated risk of CAD and MI, which was partially mediated by CAC. Lower OCN levels found in previous observational studies might be due to reverse causation.