Bioreactors have been used for bone graft engineering in pre-clinical investigations over the past 15 years. The ability of bioreactor-incubated bone marrow nuclear cells (BMNCs) to enhance bone-forming potential varies significantly, and the three-dimensional (3D) distribution of BMNCs within the scaffold is largely unknown. The aims of this study were (1) to investigate the efficacy of a carbonated hydroxyapatite (CHA) with/without BMNCs on spine fusion rate and fusion mass microarchitecture using a highly challenging two-level posterolateral spine fusion without instrumentation; and (2) to evaluate 3D distribution of BMNCs within scaffolds characterized by immunohistochemistry. Fusion rate and fusion mass were quantified by micro-CT, microarchitectural analysis, and histology. While the homogenous 3D distribution of BMNCs was not observed, BMNCs were found to migrate towards a substitute core. In the autograft group, the healing rate was 83.3%, irrespective of the presence of BMNCs. In the CHA group, also 83.3% was fused in the presence of BMNCs, and 66.7% fused without BMNCs. A significant decrease in the fusion mass porosity (p = .001) of the CHA group suggested the deposition of mineralized bone. The autograft group revealed more bone, thicker trabeculae, and better trabecular orientation but less connection compared to the CHA group. Immunohistochemistry confirmed the ability of bioreactors to incubate a large-sized substitute coated with viable BMNCs with the potential for proliferation and differentiation. These findings suggested that a bioreactor-activated substitute is comparable to autograft on spine fusion and that new functional bone regeneration could be achieved by a combination of BMNCs, biomaterials, and bioreactors.