Twenty-five rabbits were used to test the healing of diaphyseal defects by means of a reabsorbable tubular chamber. In 20 animals a 10 mm gap was created in the left radius and bridged with a poly-L-lactide chamber (experimental side). The same defect was created on the right radius, remaining untreated (control side). In the remaining 5 animals, the right radius was left intact and they were used only for biomechanical tests at 12 months. X-rays and histomorphometry showed no bone repair at the control side while the inner surface of the chamber was progressively lined by neoformed bone (75% of the normal cortical thickness). Biomechanical data demonstrated that elastic modulus and ultimate stress were not significantly different in both experimental and intact radii. The chamber, by preventing the passage of soft tissues and creating a close space for intrinsic repair factors, seems to be a suitable device for the repair of long bone defects.