The development of sufficient tissue engineered bone grafts for alveolar cleft osteoplasty could reduce the necessity of autogenous bone grafts and its donor site morbidity. The aim of the study was to evaluate tissue engineered bone grafts in an artificially created bone defect. Bone grafts were created in vitro colonizing a synthetic hydroxyapatite-tricalciumphosphate scaffold (BONITmatrix^®) with either undifferentiated mesenchymal stromal cells (group 1) or osteogenic differentiated mesenchymal stromal cells (group 2). Cells were multiplied from bone marrow of donor rats. Unmodified scaffolds (group 3) and the tissue engineered bone grafts were inserted into artificial maxillary defects of 54 Lewis rats. In 18 animals the defects remained unfilled (control). After one, three and six weeks the rats were sacrificed. The defect was evaluated radiologically and histologically with regard to the remaining defect volume and diameter. Statistical analysis followed. The bone grafts led to a specific bone formation at the defect margin. No complete reunion of any defect was observed within the healing time. After six weeks, the remaining defect volume was 6.86 ± 3.21 mm³ (control), 4.08 ± 1.36 mm³ (group 1), 5.00 ± 0.84 mm³ (group 2) 5.50 ± 1.05 mm³ (group 3). The remaining defect diameter measured 2.63 ± 0.52 mm (control), 2.39 ± 0.23 mm (group 1), 2.53 ± 0.22 mm (group 2) and 2.70 ± 0.66 mm (group 3). In all experimental groups the defect volume and diameter decreased over time, which was significant for group 1 (p = 0.014), group 2 (p = 0.025) and group 3 (p = 0.048). The defect volume and width was significantly reduced for bone grafts containing undifferentiated cells compared to control (p = 0.035) or scaffolds only (p = 0.05). Conclusion Tissue engineered bone grafts induce a pronounced bone formation in artificial bone defects compared to unfilled controls or scaffolds only.