Background New tissue engineering strategies for bone regeneration need to be investigated in a relevant preclinical large animal model before making the translation into human patients. Therefore, our interdisciplinary group established a simplified large animal screening model for intramembranous bone defect regeneration in cancellous and cortical bone. Methods Related to a well-established model of cancellous drill hole defect regeneration in sheep, both the proximal and distal epimetaphyseal regions of the femur and the humerus were used bilaterally for eight drill hole cancellous defects (Ø 6 mm, 15 mm depth). Several improvements of the surgical procedure and equipment for an easier harvest of samples were invented. For the inclusion of cortical defect regeneration, a total of eight unicortical diaphyseal drill holes (6 mm Ø) were placed in the proximal-lateral and distal-medial parts of the metacarpal (MC) and metatarsal (MT) diaphyseal bone bilaterally. Acting moments within a normal gait cycle in the musculoskeletal lower limb model were compared with the results of the biomechanical in vitro torsion test until failure to ensure a low accidental fracture risk of utilized bones (ANOVA, p < 0.05). The model was tested in vivo, using thirteen adult, female, black-face sheep (Ø 66 kg; ± 5 kg; age ≥ 2.5 years). In a two-step surgical procedure 16 drill holes were performed for the investigation of two different time points within one animal. Defects were left empty, augmented with autologous cancellous bone or soft bone graft substitutes. Results The in vitro tests confirmed this model a high comparability between drilled MC and MT bones and a high safety margin until fracture. The exclusion of one animal from the in vivo study, due to a spiral fracture of the left MC bone led to a tolerable failure rate of 8 %. Conclusions As a screening tool, promising biomaterials can be tested in this cancellous and cortical bone defect model prior to the application in a more complex treatment site.