In this study, cylindrical titanium plasma-sprayed dell implants were divided into three groups and implanted in rabbit femurs as: (1) control, a conventional method recommended by IMZ, (2) experimental I, artificial bone defects were formed prior to implantation, and (3) experimental II, porous hydroxyapatite (HA) granules were grafted into bone defects after implantation. The rabbits were sacrificed after 8 weeks and specimens were prepared and immediately pushout-tested for shear-bond strength of the bone-implant interface. Each group was examined histologically. The results showed that mean shear strength of the bone-implant interface in the control group was 2.61 ± 0.68 MPa, in experimental I was 0.66 ± 0.32 MPa, and in experimental II was 2.28 ± 0.60 MPa. There was a significant difference between the control and experimental I and between experimental I and experimental II, but no statistically significant difference between the control and experimental II. Cortical bone formation was more advanced than cancellous bone around the implants. In the histologic study of nondecalcified specimens, control and experimental II groups showed about 55% bone or osteoid tissue in contact with titanium at the bone-implant interface. In the histologic review of undecalcified specimens, the experimental I group demonstrated about 45% bone or osteoid tissue in contact with titanium at the interface, and partial bony defects were seen in the coronal zone. In the experimental II group, direct bony contact to HA granules and infiltration of a few giant cells were seen. When porous HA granules were grafted in the randomly made bone defects around the titanium implants, bone growth was more active and implant-bone bond strength was greater than when only the implants were placed. Keywords: bone-implant interface, hydroxyapatite, IMZ implants, shear-bond strength