The purpose of this study was to evaluate histomorphometrically a toothash - plaster of Paris mixture associated with collagen membrane ($Bio-Gide^{(R)}$), regarding new bone formation in the peri-implantitis defects in dogs. Three mandibular molars were removed from 1-year-old mongrel dogs. After 2 months of healing, 2 titanium implants with sandblasted with large grit and acid etched (SLA) surface were installed in each side of the mandible. Experimental peri-implantitis was induced with ligatures after successful osseointegration. Ligatures were removed after identification of bone defect beneath the level of 5th thread of fixture on radiographic image. The mucoperiosteal flaps were elevated and the contaminated fixtures were treated with chlorhexidine and saline. The bone defects were assigned to one of the following treatments: no guided bone regeneration (GBR) procedure (group 1), GBR with Bio-$Oss^{(R)}$ and Bio-$Gide^{(R)}$ (group 2), or GBR with toothash - plaster of Paris mixture (TPM) and Bio-$Gide^{(R)}$ (group 3). The dogs were sacrificed after 8 or 16 months. The mean percentages of new bone formation within the limits of the 5 most coronal threads were $17.83{pm}10.69$ (8 weeks) and $20.13{pm}13.65$ (16 weeks) in group 1, $34.25{pm}13.32$ (8 weeks) and $36.33{pm}14.21$ (16 weeks) in group 2, and $46.33{pm}18.39$ (8 weeks) and $48.00{pm}17.78$ (16 weeks) in group 3, respectively. The present study confirmed statistically considerable new bone formation within the threads in group 3 compared with group 1 at 8 and 16 weeks (P<0.05). Although, data analysis did not reveal significant differences between group 2 and 3, the latter showed better results during the period of 8 or 16 weeks. Our findings support the effectiveness of TPM as a GBR material in the treatment of peri-implantitis bone defect.