Diabetes mellitus is associated with an increased incidence of skeletal abnormalities, resulting in lower bone formation and/or remodeling. Osteopenia, osteoporosis and an increased incidence of non-traumatic fractures has been particularly observed in patients with type 2 diabetes. Recently, we have demonstrated that metformin has in vitro and in vivo osteogenic effects: (a) it stimulates the proliferation, differentiation and mineralization of osteoblasts in culture, and (b) in non-diabetic rats, it increases the repair of minimal bone lesions and improves femoral trabecular bone microarchitecture. In this study, we evaluated in rats the effect of diabetes induction by a partial destruction of pancreatic beta cells, on the osteogenic commitment of bone marrow progenitor cells (BMPC), and the modulation of this effect by orally administered metformin. We used young male Sprague Dawley rats (200 g), divided into four groups: untreated non-diabetic controls [C], non-diabetic rats treated for 2 weeks with metformin administered in drinking water (100 mg/kg/day) [M], untreated diabetic rats [D], and diabetic rats treated for 2 weeks with metformin (100 mg/kg/day) [DM]. Induction of Diabetes was performed one week prior to treatment with metformin, by successive intraperitoneal injections with 75 mg/kg body weight of nicotinic acid and 60 mg/kg body weight of streptozotocin. At the end of all treatments, blood samples were obtained to confirm the development of Diabetes, after which the animals were sacrificed by cervical dislocation under anesthesia. Femora and/or tibiae were dissected, and bone marrow cells were collected by flushing the bone diaphysal canal with Dulbecco’s modified essential medium (DMEM) under sterile conditions. Adherent cells were grown to confluence in DMEM-10 % fetal bovine serum (FBS), after which we assessed alkaline phosphatase specific activity (ALP) by an enzymatic kinetic method, and type 1 collagen production (Col-1) by a Sirius Red colorimetric method (basal osteoblastic differentiation of BMPC). Subsequently, BMPC were submitted to an osteogenic induction for 15 days with an osteogenic medium (DMEM-10 % FBS containing ascorbic acid and sodium beta-glycerophosphate), after which ALP and Col-1 were evaluated. Basal ALP activity and type 1 collagen production (BMPC without osteogenic differentiation) showed no significant differences between the four experimental groups. After 15 days of culture in osteogenic medium, BMPC from control rats increased their expression of ALP (5 times compared to baseline) and collagen production (11 times compared to baseline). BMPC from diabetic rats after 15 days culture in osteogenic medium, also showed a significant (although smaller) increase in ALP (2-3 fold over basal activity) and collagen production (4-fold compared to baseline). BMPC obtained from rats treated with metformin (groups M and DM) and submitted to osteogenic induction for 15 days, showed an approximately 2-4-fold increase in both ALP and Col-1 (when compared with groups C and D, respectively). After 21 days of osteogenic induction, a decrease was observed in the mineralization of BMPC obtained from group D (65 % of that for group C). Treatment with metformin increased the mineralizing capacity of BMPC in all cases, including a reversal of the inhibitory effect of Diabetes on this parameter. In conclusion, we have found that our model of Diabetes reduces the osteogenic potential of bone marrow progenitor cells, and that this effect is partially reverted by orally administered metformin. These findings could explain, at least in part, the bone alterations that have been associated with Diabetes mellitus.