Objective Based on multi-objective optimization, a design method for Voronoi bionic porous scaffolds tailored to different extents of bone defects was proposed. Methods First, the effects of design parameters on mechanical and biological properties of the scaffolds were investigated. The response surface models were then established respectively for the design parameters and performance indicators (specific surface area, elastic modulus, yield strength, and permeability). Using a cubic scaffold with side length of 15 mm as an example (assuming a corresponding bone defect of the same dimension), multi-objective optimization of the scaffold was finally conducted using the non-dominated genetic algorithm-II algorithm, while considering the elastic modulus and permeability ranges of bone tissues as performance constraints. Results The degree of anisotropy in Voronoi scaffolds was influenced by the number of seed points, while the size and scaling factors of the scaffolds exclusively impacted the rod diameter and rod length. Using the design method of this study, the optimal scaffold with specific defect size satisfying mechanical and biological properties was obtained. The optimal scaffold under different strength requirements was obtained by adjusting the yield strength to change the utopia point. Conclusions A design method for Voronoi bionic porous scaffolds based on multi-objective optimization was proposed. This method can be applied to bone defects at different degrees and provides a new idea for the personalized design of bone tissue engineering scaffolds.