Five titanium plasma-sprayed implants were biopsied from a human volunteer 6 months after placement. Four test implants had been placed in immediate extraction sockets, while one implant was placed in a mature site and served as a control. The histologic analysis demonstrated that all five implants achieved osseointegration as demonstrated by light microscopy, whereas a varying degree of bone-implant contact was observed. The non-loaded control implant had the highest percentage of bone-implant contact, 72%, followed by the two implants placed in the canine sites presenting with a horizontal defect dimension of 1.5 nm or less. These implants were placed without a barrier membrane, but in a submerged fashion. The histometric analysis showed a mean bone-implant contact of 50% for these two implants. The lowest mean bone-implant contact (17%) was observed for the two molar implants, which had horizontal defect dimensions of 4 mm; these implants were placed in a non-submerged fashion with the implants perforating an expanded polytetrafluoroethylene membrane. The authors concluded that osseointegration may occur in immediate extraction sites in humans using titanium implants with a plasma-sprayed surface. The horizontal component of the peri-implant defect was apparently the most critical factor relating to the final amount of bone-implant contact.