Purpose: To compare bone substitutes composed of glycated collagen with synthetic micro-sized (1 to 10 μm) hydroxyapatite (OB) vs non–cross-linked collagen matrix with large-particle (250 to 1,000 μm) bovine-derived hydroxyapatite (BOC). Materials and Methods: The P1 to P4 premolars were bilaterally extracted from the mandibles of 19 Beagle dogs. After 21 days, osteotomies were created in each dog that received OB or BOC and were covered with a collagen membrane or were left untreated. The animals were randomly divided into three groups based on sacrifice time (4, 12, or 24 weeks). The right and left hemimandibles were trimmed to facilitate imaging and histology, and all tissues were placed in 10% neutral-buffered formalin. Microcomputed tomography (MicroCT 40 Scanner, Scanco) was used to analyze bone sections. Bone volume, residual material volume, and bone mineral density were determined for each treatment site (OB and BOC) based on a volume of interest that encompassed the original defect. Additionally, blinded histopathologic assessment (based on the ISO 10993-6 scoring system) and histomorphometry were performed on sections ground to < 100 μm thick and stained with Stevenel’s blue. Results: No clinical side effects were noted. No statistical differences were observed for OB vs BOC regarding the mineral volume percentage. Compared to OB, BOC had significantly higher mean mineralization densities at 12 weeks (P < .01), but this difference did not extend to 24 weeks. For residual grafting material, bone maturation, alveolar ridge restoration, and inflammatory response, OB showed a residual amount of bone graft and no statistical differences compared to BOC. Conclusion: Both OB and BOC represent valid treatment options for critically sized bone defects. Both bone fillers outperformed the sham-operated, ungrafted (empty) control, demonstrating statistically improved bone growth and ridge restoration.