The purpose of this study was to determine which treatment of a large osseous defect adjacent to an endosseous dental implant would produce the greatest regeneration of bone and degree of osseointegration: barrier membrane therapy plus demineralized freeze-dried bone allograft (DFDBA), membrane therapy alone, or no treatment. The current study histologically assessed changes in bone within the healed peri-implant osseous defect. In a split-mouth design, 6 implants were placed in edentulous mandibular ridges of 10 mongrel dogs after preparation of 6 cylindrical mid-crestal defects, 5 mm in depth, and 9.525 mm in diameter. An implant site was then prepared in the center of each defect to a depth of 5 mm beyond the apical extent of the defect. One mandibular quadrant received three commercially pure titanium (Ti) screw implants (3.75 x 10 mm), while the contralateral side received three hydroxyapatite (HA) coated root-form implants (3.3 x 10 mm). Consequently, the coronal 5 mm of each implant was surrounded by a circumferential defect approximately 3 mm wide and 5 mm deep. The three dental implants in each quadrant received either DFDBA (canine source) and an expanded polytetrafluoroethylene membrane (ePTFE), ePTFE membrane alone, or no treatment which served as the control. Clinically, the greatest increase in ridge height and width was seen with DFDBA/ePTFE. Histologically, statistically significant differences in defect osseointegration were seen between treatment groups (P < 0.0001: DFDBA/ePTFE > ePTFE alone > control). HA-coated implants had significantly greater osseointegration within the defect than Ti implants (P < 0.0001). Average trabeculation of newly formed bone in the defect after healing was significantly greater for HA-coated implants than for titanium (P < 0.0001), while the effect on trabeculation between treatments was not significantly different (P = 0.14). Finally, there were significantly less residual allograft particles in defect areas adjacent to HA-coated implants than Ti implants (P = 0.0355). The use of HA-coated implants in large size defects with DFDBA and ePTFE membranes produced significantly more osseointegration histologically than other treatment options and more than Ti implants with the same treatment combinations. The results of this study indicate that, although the implants appeared osseointegrated clinically after 4 months of healing, histologic data suggest that selection of both the implant type and the treatment modality is important in obtaining optimum osseointegration in large size defects.