From a biological standpoint, the best material for reconstruction of bone defects is the autogenous bone graft. However, as tissue healing is affected under diabetic conditions, major changes might take place in the revascularization, incorporation, replacement and remodeling phases of the grafted area. The purpose of this study was to assess the bone healing process in surgical wounds prepared in tibiae of diabetic rats and filled with autogenous bone. Forty male rats (Rattus norvegicus albinus, Wistar) were randomly assigned to receive an endovenous injection (penile vein) of either citrate buffer solution (Group 1 - control; n=20) or streptozotocin dissolved in citrate buffer solution (35 mg/kg) to induce diabetes (Group 2 - diabetic; n=20). After determination of glycemia, the animals were anesthetized and the anterolateral regions of the tibiae of both limbs were shaved, antisepsis was performed and longitudinal incisions were made in each limb. The tibiae were exposed and two 2mm-diameter surgical cavities were prepared: one in the right limb, filled with particulate autogenous bone and the other in the left limb, filled with blood clot. The animals were euthanized at 10 and 30 postoperative days. The anatomic pieces were obtained, submitted to laboratory processing and sections were stained by hematoxylin and eosin and Masson's Trichrome for histomorphologic and histometric analyses. In both groups, the wounds filled with autogenous bone graft showed better results than those filled with blood clot. The control group showed higher new bone formation in wounds filled with autogenous bone graft at 30 days than the diabetic group, but without statistical significance. It may be concluded that, in general, the new bone formation occurred with autogenous graft was quantitatively similar between control and diabetic groups and qualitatively better in the control group.