Computational approaches have great potential for aiding clinical product development by finding promising candidate designs prior to expensive testing and clinical trials. Here, an approach for designing multilevel bone tissue scaffolds that provide structural support during tissue regeneration is developed by considering mechanical and biological perspectives. Three key scaffold design properties are considered: (1) porosity, which influences potential tissue growth volume and nutrient transport, (2) surface area, which influences biodegradable scaffold dissolution rate and initial cell attachment, and (3) elastic modulus, which influences scaffold deformation under load and, therefore, tissue stimulation. Four scaffold topology types are generated by patterning beam or truss-based unit cells continuously or hierarchically and tuning the element diameter, unit cell length, and number of unit cells. Parametric comparisons suggest that structures with truss-based scaffolds have higher surface areas but lower elastic moduli for a given porosity in comparison to beam-based scaffolds. Hierarchical scaffolds possess a large central pore that increases porosity but lowers elastic moduli and surface area. Scaffold samples of all topology types are 3D printed with dimensions suitable for scientific testing. A hierarchical scaffold is fabricated with dimensions and properties relevant for a spinal interbody fusion cage with a maximized surface-volume ratio, which illustrates a potentially high performing design configured for mechanical and biological factors. These findings demonstrate the merit in using multidisciplinary and computational approaches as a foundation of tissue scaffold development for regenerative medicine.