The purpose of this study was to compare extraction socket healing in 8 patients after implantation with either xenogenic bovine bone (n = 5 sites), demineralized freeze-dried bone (DFDBA) (n = 3 sites), autologous bone (n = 3 sites), or human bone morphogenetic proteins in an osteocalcein/osteonectin carrier (hBMP/NCP) (n = 2 sites). Three of the patients received 6 commercially pure micro screws which were fixed into extraction sockets, after which the sockets were implanted with either bovine bone (n = 3 sites), DFDBA (n = 2 sites) or intraoral autologous bone (n = 1 site). Biopsies of the extraction sockets were taken from 3 to 6 months after treatment (average, 4.6 months). For comparison of healing between the implanted materials, histologic evaluation and bone scores were determined. Bone scores of 0 indicated an absence of new bone, with dead implanted bone particles entrapped within connective tissue, while a score of 3 indicated the entire field consisted of vital bone. Biopsies from bovine bone sockets revealed dead implanted particles surrounded by connective tissue. Isolated sections showed host bone in contact with the bovine bone particles. Bone scores ranged from 0 to 3. Biopsies from DFDBA-implanted sites revealed dead particles entrapped with dense connective tissue. The bone scores ranged from 0 to 1. Biopsies from sites implanted with hBMP/NCP revealed a combination of woven and lamellar bone with bone scores of 3. Five of the 6 micro screws were processed and evaluated. One screw was mobile at the time of removal and was not evaluated. Bone scores were used to compare new bone formation adjacent to the micro screws. Bone scores ranged from 0 to 2. A score of 0 indicated non-vital implant material in contact with host bone and connective tissue in contact with implant; 2 indicated vital bone in contact with the majority of the implant surface. Retrieved sockets with micro screws implanted with bovine bone (n = 2) demonstrated a connective tissue interface between the screws and the surrounding tissues (bone score 0). The adjacent tissues showed dead bovine particles entrapped within fibrous tissue. Retrieved screws implanted with DFDBA (n = 2) were surrounded by connective tissue, with dead bone particles enmeshed within fibrous tissue (bone score 0). The screw implanted with intra-oral autologous bone was primarily surrounded by vital bone with a connective tissue interface (bone score 1). Three implant threads were in contact with bone. The results of this study indicate that bovine bone, DFDBA, and intraoral autologous bone do not promote extraction socket healing. Sockets implanted with hBMP/NCP contained vital woven and lamellar bone. Xenogenic bovine bone and DFDBA did not contribute to bone to micro screw contacts and are not recommended for enhancement of vital bone to implant contacts. Intraoral autogenous bone also does not appear to significantly contribute to bone to implant contacts. Intraoral autologous bone, xenogenic bone, and DFDBA appear to interfere with normal extraction socket healing. J Periodontol 1998;69:414–421.