In this study of six greyhound mandibles, 24 implants were placed into extraction sockets. Eighteen of the implants were connected with abutments and immediately placed into normal function, and the remaining six submerged implants serv ed as controls. At placement, dehiscence areas were created adjacent to 12 of the implants to study bone regeneration in extended membrane-portected defects. Six of the defects were covered with expanded polytetrafluooethylene membranes, and six of the defects were augmented with autologous bone grafts and covered with expanded polytetrafluoroethylene material. The clinical and radiographic evaluation demonstrated that all implants achieved functional osseointegration and no implants were lost. Implant sites where membranes were used to regenerate bone over dehiscence defects demonstrated a high rate of membrane exposure and variable response to treatment. On completion of the study at 16 weeks, histomorphometric analysis demonstrated osseointegration with direct bone-to-implant contact for all implants, although the percentage of contact varied considerably. Histomorphometry indicates that the quality of osseointegration may be less favorable than the clinical and radiographic result would suggest.