An osseous defect was created in the acetabulum and femur of twenty dogs, and then a fiber-metal total hip prosthesis was inserted. A comparison was made between the effects of leaving the defects unfilled, filling the defects with an autogenous bone graft, and filling them with a 50:50 mixture of autogenous bone graft and a biphasic ceramic composed of hydroxyapatite and tricalcium phosphate. The characteristics of formation of bone within the sites of the defects and the extent of the ingrowth of bone into the underlying porous surface were analyzed. At six and twelve weeks after implantation, the dogs in the control group (unfilled defects) had the least amount of bone in the sites of the defects. In the dogs that had had an autogenous bone graft and those in which the defects had been filled with a mixture of autogenous bone graft and biphasic ceramic, there were excellent osteoconductive properties in the filling of the sites of the defects with new bone. A comparison was made between the amount of bone that formed in the osseous defects and the amount that formed in the spaces of the porous-surfaced fiber-metal components of the prostheses, directly underlying and adjacent to the defects. At six weeks, the greatest amount of ingrowth of bone into the spaces of the underlying titanium fiber-metal acetabular components was seen in the control group (22 per cent of the porous surface), followed by the group in which the defects had been filled with a mixture of autogenous bone graft and biphasic ceramic (19 per cent). The defects that had been filled only with autogenous bone graft had poor ingrowth of bone into the porous surface (4 per cent) and predominantly fibrous ingrowth. At twelve weeks, these defects again showed the least amount of ingrowth of bone (15 per cent) compared with the defects in the control group (24 per cent) and those that had been filled with a mixture of autogenous bone graft and biphasic ceramic (24 per cent). No statistically significant differences were found in the amount of ingrowth of bone into the defects that had been created in the femur in the three groups of dogs at either six or twelve weeks. However, the trends appeared to correlate with the data for the acetabular defects.