The purpose of this study was to determine the optimal ratio of calcium hydroxyapatite (HA) to beta tricalcium phosphate (beta TCP) in a biphasic porous calcium phosphate (BCP) ceramic for effective repair of periodontal osseous defects. Defects were surgically produced in beagle dogs and made chronic for 4 months to simulate periodontal disease. Mucoperiosteal periodontal flaps were reflected, followed by osseous defect debridement and root planing. Specially prepared ceramic with different HA/beta TCP ratios were implanted into the prepared defects. The sites were allowed to heal for 6 months, animals were euthanized, and site-blocks were removed for histological study. During the follow-up phase, scaling and polishing were done once a month, and standardized probing attachment levels were recorded pre- and 6-months postoperatively. The Duncan's multiple range test showed that all the treatments produced statistically significant higher gain in probing attachment levels than the control group (0HA/0 beta TCP) (P < 0.05). Among the 7 "active" treatment groups, 2 (65/35 and 85/15) had significantly higher gain in probing attachment levels than those in 3 groups (50/50, 100/0, and 0/100) (P < 0.05). Histologically, higher HA ratio (but not 100% HA) showed accelerated new bone formation and new attachment levels. Based on histological results, the 85HA/15 beta TCP ratio appears to demonstrate greater gain in attachment level and bone regeneration in the treatment of periodontal osseous defects.