Objective Most existing studies on the effect of bone marrow on bone mechanical properties have been based on trabecular bone models reconstructed from high-resolution CT images, and few have addressed the impact of bone marrow fluid on the strength of the entire vertebral body. This study aims to investigate the effect of bone marrow fluid in the vertebral body on vertebral strength and fracture risk assessment using QCT-based finite element modeling analysis. Methods First, 30 subjects underwent thoracolumbar spine MRI and QCT scans and were divided into fracture and non-fracture groups, from which vertebral bone marrow adipose tissue (BMAT) content and volumetric BMD (vBMD) were measured. Then, poroelastic-plastic models (incorporating bone marrow) and elastic-plastic models (excluding bone marrow) were established to calculate vertebral strength. Finally, Wilcoxon signed-rank test, independent-samples t-tests and receiver operating characteristic (ROC) analysis were used to evaluate the effect of bone marrow on vertebral strength and fracture risk assessment. Results Wilcoxon signed-rank test showed that when bone marrow fluid in cancellous bone was considered, vertebral strength significantly increased by 3%, 7%, and 1% in all subjects, the fracture group, and the control group, respectively. Independent-samples t-tests indicated significant differences in vertebral strength between the fracture and control groups, regardless of whether bone marrow fluid was considered (37% and 32% higher than those in the fracture group, respectively). ROC analysis showed no difference in the AUC values of vertebral strength before and after considering bone marrow fluid, but both were higher than that of vBMD (0.78, 0.79, and 0.63, respectively). Conclusion The presence of bone marrow enhances vertebral strength to a certain extent. Moreover, regardless of whether the mechanical effect of bone marrow is considered, bone strength outperforms BMD in assessing vertebral fracture risk.