This study was performed to evaluate the effect of deproteinized bovine porous bone mineral (BBM) and BBM-collagen (BBMC) used alone or in combination with a bilayer collagen membrane in guided periodontal regeneration. In 12 dogs, contralateral surgical circular fenestration defects 5 mm in diameter were produced at the midbuccal aspect of the alveolar bone in 24 maxillary canines. Bone, periodontal ligament, and cementum were completely removed. Experimental sites were filled with BBM or BBMC. Bilayered collagen membranes covered half the experimental sites (BBM+M and BBMC+M), and the other half were left uncovered. Control sites remained empty; half were covered with collagen membranes (cont+M) and the underlying space spontaneously filled with blood, and half were left uncovered (cont). Three months postsurgery, undecalcified sections were prepared. Measurements were made using a caliper on a projection microscope, and the surface area of new bone and BBM particles within the healed surgical defect was evaluated using the point-counting method. In the experimental defects, new cementumcovered 31% to 67% of the exposed dentin, with a significant difference between defects covered with membranes and defects that were not covered (P .05). New cementum in the control (unfilled) defects also differed significantly between covered and uncovered defects. New bone growth presented a pattern similar to the cementum. There was no statistical difference between defects treated with BBM and BBMC, within both covered and uncovered groups. There was less connective tissue in the covered defects than in the uncovered defects (P .05). The defects were filled with new bone, new connective tissue/bone marrow, and bovine bone particles. New bone area fraction was 23.4% to 25.2% in defects filled with BBMC and BBM, respectively (P = NS). Bone fraction area in membranecovered defects ranged from 34.4% to 36.8% in experimental defects (P = NS). All membrane-treated defects showed higher values for bone area fraction in comparison to the uncovered control defects. Particle area fraction ranged between 17.4% and 26.2%, with only BBMC and BBM+Mdefects showing a statistically significant difference (P .05). Defects filled with submembranous blood clot exhibited significantly more new cementumand bone regeneration than experimental defects filled with BBM or BBMC. Treatment of defects with BBM or BBMC showed similar influences on bone and cementum regeneration in fenestration periodontal defects. The presence or absence of bilayered collagen membranes was the predominant factor influencing bone and cementum regeneration.