Aim: The aim of this pre-clinical in vivo study was to analyse different stages of wound healing after guided bone regeneration in non-contained mandibular buccal bone defects. Materials and methods: Eighteen female beagle dogs, between 1.5 and 2 years old, were used. Buccal bone defects were created in the mandible following extraction of the mesial roots of M1, P4, the distal root of P3 and booth roots of P2. Augmentation procedures of the healed defects were performed 3 months later using a bone replacement graft (T1), an absorbable collagen membrane (T2) or a combination of both procedures (T3). Using a randomized block study design, four stages of healing in two groups of dogs were examined (4 days, 2, 6 weeks and 3 months). The animals were euthanized, and biopsies obtained at the end of each of the study periods were prepared for histological examination. Results: The different reconstructive procedures resulted in regenerated tissue compartments of varying size that contained newly formed bone, non-mineralized tissue and bone augmentation biomaterial when a bone replacement graft was used. While the proportions of mineralized tissue increased and non-mineralized tissue decreased over time in the three groups, the changes in proportions of the DBBM material were small. Initial defect depth, healing time and treatment group significantly influenced the percentage of mineralized tissue obtained. The multivariate multilevel analysis showed that significantly larger area proportions of mineralized tissue were obtained when the T2 sites were compared with T1 and T3 sites, what highlights the importance of the barrier membrane effect for attaining new bone formation. Only in the larger size defects (M1) total ROI at T3 and T1 sites was significantly larger than at T2, what highlights the importance of using a bone replacement graft as a space maintenance scaffold. Conclusion: It is suggested that healing following augmentation of non-contained buccal bone defects was characterized by a gradual shift in the relative proportions of non-mineralized and mineralized tissue components. © 2017 John Wiley & Sons A/S.