To evaluate the preclinical efficacy and safety of human mesenchymal stem cells (hMSC) rapidly expanded in growth medium for clinical use with human serum and recombinant growth factors, we conducted a controlled, randomized trial of plasma clots with hMSC vs. plasma clots only in critical segmental femoral defects in rnu/rnu immunodeficient rats. X-ray, microCT and histomorphometrical evaluation were performed at 8 and 16 weeks. MSC were obtained from healthy volunteers and patients with lymphoid malignancy. Human MSC survived in the defect for the entire duration of the trial. MSC from healthy volunteers, in contrast to hMSC from cancer patients, significantly improved bone healing at 8, but not 16 weeks. However, at 16 weeks, hMSC significantly improved vasculogenesis in residual defect. We conclude that hMSC from healthy donors significantly contributed to the healing of bone defects at 8 weeks and to the vascularisation of residual connective tissue for up to 16 weeks. We found the administration of hMSC to be safe, as no adverse reaction to human cells at the site of implantation and no evidence of migration of hMSC to distant organs was detected.