Bilateral closed femoral fractures were produced in two groups of rats. Reaming was made from the trochanteric area before fracture, and the fractures were stabilized by intramedullary pinning. In one (hematoma) group, both femurs were exposed subperiosteally at the midshaft prior to fracture. At one side, the hematoma was evacuated 30 min following fracture, at the other side the hematoma was left undisturbed. In the other (periosteal) group, the femur was exposed subperiosteally at one side, while the periosteum was left intact at the other side. The healing of the fractures was evaluated at 4 weeks. In the hematoma group, no differences were found in callus production, while bending moment and bending rigidity were greater at the side where the hematoma was not removed. No differences were found in fracture energy. In the periosteal group, marginal differences were found in the callus area. Bending moment, bending rigidity and fracture energy were greater at the side where the periosteum was left intact. A comparison of all four limbs, showed that the periosteal sealing was of particular importance for rapid healing.