Background <p>In observational studies, type 2 diabetes (T2D) and the risk of osteomyelitis are related, but the causality of the association has not been fully established. We conducted a Mendelian randomization study to assess the causal relationship.</p> Methods <p>A significant threshold (<i>p</i> &lt; 5 × 10^− 8) was used for exposure-related genetic instrument screening in a large-scale genetic summary data from genome-wide association (GWAS) studies. Univariate and multivariable mendelian randomization (MR) analyses were used to evaluate the causal relationship. The causal analysis methods mainly used the inverse variance weighted (IVW) random effects method, and the weighted median, MR-Egger method were used for sensitivity analysis.</p> Results <p>In univariate MR analysis, IVW estimates showed that higher BMI (OR = 1.524; 95% CI 0.255–1.851; <i>p</i> = 2.16 × 10 − 5), Waist-Hip ratio (adjusted for BMl) (OR = 1.906; 95% CI 1.28–2.887; <i>p</i> = 2 × 10^− 3), and T2D (OR = 1.324; 95% CI 1.206–1.454; <i>p</i> = 3.66 × 10^− 9) were associated with an increased risk of osteomyelitis. Higher economic income (OR = 0.332; 95% CI 10.195–0.567; <i>p</i> = 5.24 × 10^− 5) reduced the risk of osteomyelitis. In multivariable MR, the association between T2D and osteomyelitis continued to exist after mutual adjustment (OR = 1.264; 95% CI 1.122–1.423; <i>p</i> = 1.11 × 10^− 4).</p> Conclusion <p>This study supports the independent causative role of T2D in osteomyelitis.</p>