Aim of the study. To estimate and compare the cost implications of the first attempt of treatment of tibial fracture non-unions treated with either autologous bone grafting or BMP-7. Materials and methods. Patients who were successfully treated for fracture non-unions between 2001 and 2005 were included. Exclusion criteria included infected non-union, children, malignancy, or chronic debilitating disease. The decision to use BMP-7 or autogenous bone graft was guided by the defect size and the surgeon's preferred method of treatment. Group 1 (n = 12) received