The midplate structural rigidities of metal- and plastic-plated intact canine femora were experimentally determined after initial plate application and 8 weeks after in vivo implantation. Composite beam theory significantly overestimated the bending rigidities of the metal-plated bones. The rigidities of the plastic-plated bones were nearly identical to that of the isolated bone, as composite beam theory predicted. Plating with either plate increased the intracortical porosity and caused the deposition of periosteal new bone, which was greater with plastic than with metal plates. The increased rigidities provided by the attachment of the metal plates and, hence, the degree of bone strain shielding were variable. Platings for 8 weeks which provided little strain shielding with either metal or plastic plates caused an increase in bone flexural rigidity (measured after plate removal) with respect to the contralateral control. Platings that provided increasing amounts of strain shielding caused a decreasing midplate bone rigidity (measured after plate removal) and increasing bone deposition at the outer screws. These findings suggest that the surgical implantation of any plate (metal or plastic) will provide a net stimulus to bone formation and consequently increased bone structural rigidity, even though intracortical porosity is increased. If the plate significantly reduces the normal loads borne by the bone, however, there is a net stimulus to remove bone, resulting in a loss of midplate structural rigidity within 8 weeks of implantation.