Limitations in the use of autologous bone graft, which is the gold standard therapy in bone defect healing, drive the search for alternative treatments. In this study the influence of rhTGFb-3 on mechanical and radiological parameters of a healing bone defect in the sheep tibia was assessed. In the sheep, an 18-mm long osteoperiosteal defect in the tibia was treated by rhTGFb-3 seeded on a poly(L/DL-lactide) carrier (n ¼ 4). In a second group (n ¼ 4), the defect was treated by the carrier only, in a third group (n ¼ 4) by autologous cancellous bone graft, and in a fourth group (n ¼ 2) the defect remained blank. The healing process of the defect was assessed by weekly in vivo stiffness measurements and radiology as well as by quantitative computed tomographic assessment of bone mineral density (BMD) every 4 weeks. The duration of the experiment was 12 weeks under loading conditions. In the bone graft group, a marginally significant higher increase in stiffness was observed than in the PLA/rhTGFb-3 group ( p ¼ 0.06) and a significantly higher increase than in the PLA-only group ( p ¼ 0.03). The radiographic as well as the computed tomographic evaluation yielded significant differences between the groups ( p ¼ 0.03), indicating the bone graft treatment (bone/per area, 83%; BMD, 0.57 g/cm 3 ) performing better than the PLA/rhTGFb-3 (38%; 0.23 g/cm 3 ) and the PLA-only treatment (2.5%; 0.09 g/cm 3 ), respectively. Regarding the mechanical and radiological parameters assessed in this study, we conclude that rhTGFb-3 has a promoting effect on bone regeneration. However, under the conditions of this study, this effect does not reach the potential of autologous cancellous bone graft transplantation. ß