Individuals with transfemoral amputation commonly develop chronic health problems due to decreased physical activity as a result of the missing musculature and tissue on the amputated side, and the poor imitation of the intact limb provided by the prosthesis. In addition, the indirect and semi-rigid connection of the socket to the body may increase interlimb asymmetries, as well as lead to pain and discomfort on the residual limb. Recent innovations have introduced a bone-anchored or osseointegrated (OI) implant which connects the prosthesis to the skeleton, and removes most of the socket related pain and discomfort complaints, as well as providing a rigid connection which may reduce the interlimb asymmetries. However, the direct bone and prosthesis connection may also introduce longitudinal bone health concerns due to the repetitive loads during walking. This dissertation investigated the effect of walking speed on the loads placed on the lower limbs of 11 individuals who use an OI prosthesis at 3 different anatomical levels, including the whole limb through interlimb ground reaction force, the joints through interlimb joint kinematics and kinetics, and finally the residual limb bone through implant input forces, finite element analysis of bone strain, and the probability of bone injury with a simulated lifetime of use.In study 1, the interlimb ground reaction force asymmetries were found to be moderate to large at all walking speeds, and to have a general increase as individuals walked faster, indicating there is an intact limb reliance strategy which may be used to compensate for the limitations of the amputated limb. Similarly, in study 2, the interlimb joint kinematics and kinetics were found to have moderate to large asymmetries at each joint level, with a general increase in asymmetry at faster walking, with this increase largely due to limitations within the prothesis. In study 3, the abutment force decreased in magnitude with walking speed, but the peak strain on the bone, and the probability of injury was greater for the preferred speed and fast speed walking when compared to slow speed walking. However, the overall probability of injury was low for all speeds, indicating the ability of the bone to repair and adapt with sustained loading likely provides effective protection over a lifetime of simulated OI prothesis use. The findings of this dissertation suggest that the more rigid connection afforded by the OI implant cannot fully remove the interlimb asymmetries which occur as a result of the poor imitation of the intact limb provided by the prosthesis and prosthesis components, but that there is minimal risk to the bone due to a lifetime of sustained walking with an OI prosthesis as a result the inherent ability of the bone to repair and adapt to variable loads over time. Therefore, while an OI prosthesis may not fully mitigate the interlimb asymmetries which occur as a result of the prosthesis limitations, individuals who use an OI prosthesis may feel confident that there is minimal longitudinal risk to the bone as a result of walking over their lifetime.