Abstract Autosomal recessive osteopetrosis (ARO) is a severe bone disease characterized by increased bone density due to impairment in osteoclast resorptive function or differentiation. Hematopoietic stem cell transplantation is the only available treatment; however, this therapy is not effective in RANKL-dependent ARO, since in bone this gene is mainly expressed by cells of mesenchymal origin. Of note, whether lack of RANKL production might cause a defect also in the bone marrow (BM) stromal compartment, possibly contributing to the pathology, is unknown. To verify this possibility, we generated and characterized BM mesenchymal stromal cell (BM-MSC) lines from wild type and Rankl−/− mice, and found that Rankl−/− BM-MSCs displayed reduced clonogenicity and osteogenic capacity. The differentiation defect was significantly improved by lentiviral transduction of Rankl−/− BM-MSCs with a vector stably expressing human soluble RANKL (hsRANKL). Expression of Rankl receptor, Rank, on the cytoplasmic membrane of BM-MSCs pointed to the existence of an autocrine loop possibly activated by the secreted cytokine. Based on the close resemblance of RANKL-defective osteopetrosis in humans and mice, we expect that our results are also relevant for RANKL-dependent ARO patients. Data obtained in vitro after transduction with a lentiviral vector expressing hsRANKL would suggest that restoration of RANKL production might not only rescue the defective osteoclastogenesis of this ARO form, but also improve a less obvious defect in the osteoblast lineage, thus possibly achieving higher benefit for the patients, when the approach is translated to clinics.