In this study, we simulated a wide cortex separation from a cementless hip prosthesis using the bone resorption remodeling method that is based on the generation of high compressive stress around the distal cortical bone. Thereafter, we estimated the effect on late migration quantities of the hip prosthesis produced by the interface state arising from bone ingrowth. This was accomplished using cortical bone remodeling over a long period of time. Two-dimensional natural hip and implanted hip FEM models were constructed with each of the following interface statements between the bone and prosthesis: (1) non-fixation, (2) proximal 1/3, (3) proximal 2/3 and (4) full-fixation. The fixation interfaces in the fully and partially porous coated regions were rigidly fixed by bony ingrowth. The non-fixation model was constructed as a critical situation, with the fibrous or bony tissue not integrated at all into the implant surface. The daily load history was generated using the three loading cases of a one-legged stance as well as abduction and adduction motions. With the natural hip and one-legged stance, the peak compressive principal stresses were found to be under the criteria value for causing bone resorption, while no implant movement occurred. The migration magnitude of the stem of the proximal 1/3 fixation model with adduction motion was much higher, reaching 6%, 11%and 21%greater than those of the non-fixation, proximal 2/3 fixation and all-fixation models, respectively. The full-fixation model showed the lowest compressive principal stress and implant movement. Thus, we concluded that the late loosening and subsequent movement of the stem in the long term could be estimated with the cortical bone remodeling method based on a high compressive stress at the bone-implant interface. The change caused at the bone-prosthesis interface by bony or fibrous tissue ingrowth constituted the major factor in determining the extent of cortical bone resorption occurring with clinical loosening and subsequent implant movement.