Hip replacement in a femur with proximal bone loss was studied in a canine model. Specifically, the fixation stability of a prosthesis and proximal bone graft relative to the distal femur was analyzed. Linear and angular relative displacements of proximal femoral canine grafts were determined after four months of implantation and after ex vivo (nonresected) implantation. Relative displacements representing the six degrees of freedom were calculated via matrix transformations from the relative displacements measured by nine eddy current transducers. A comparison of the stability during axial and transverse loading between distally cemented and distally press-fit total hip reconstructions was made. The results showed no significant differences between the stability of cemented and cementless reconstructed femora during both loading orientations. The femora implanted for four months had significantly larger linear lateral displacement for cemented (p = 0.0490) and cementless (p = 0.0313) groups during axial loading compared to the nonresected femora. Transverse loading revealed no significant differences between control and four month groups.