Experimental characterisation of ex vivo specimens is limited by specimen availability and high costs, whereas 3D printing provides a cost-effective alternative for producing multiple replicas. This study aimed to develop a methodology for evaluating the individual and combined effects of material composition and geometry on the biomechanical performance of 3D-printed vertebrae. CT scans of healthy human vertebrae and with lytic metastases were segmented to fabricate synthetic models through Digital Anatomy Printing. Three types of 3D-printed models were produced: Healthy vertebrae, Metastatic vertebrae, and Healed vertebrae (metastatic geometry filled with healthy material). All models were tested under axial compression to measure the strength, stiffness, and strain. Repeatability across replicas was assessed as well as comparison of mechanical properties among the different vertebral types. Results showed excellent repeatability, with coefficients of variation below 5% for strength and stiffness-related parameters. The Metastatic models exhibited significant reductions in strength compared to Healthy ones, while stiffness remained similar, consistent with ex vivo data trends. Healed models highlighted the role of material composition in driving mechanical behaviour, independently of geometry. This work provides the first quantitative assessment of 3D-printed vertebrae with metastatic lesions, supporting their future potential as standardised alternatives to cadaveric testing.