Several questions relating to the biomechanics of the AO unreamed tibial nail were addressed in this study. These included the effects of the location of the nail bend on the reduction of a high proximal fracture, and the relation of proximal locking screw hole orientation and fracture component cortical contact to the mechanical stiffness of the construct. To measure fracture site malalignment with nail insertion, a motion transducer mounted on the distal tibial shaft was used to track the position of the proximal component during and after insertion of the nail. For studying the effect of screw hole orientation, the intramedullary (IM) nail was modified by drilling a second set of proximal screw holes with oblique instead of parallel alignment. The axial, torsional, and varus-valgus stiffnesses of the constructs with parallel or oblique screws and with or without fracture site contact were determined. In an experimental model with an osteotomy located proximal to the position of the bend in the nail when fully inserted, anterior displacement of the proximal fracture component (or posterior displacement of the distal component) of up to 1 cm was measured. Oblique proximal locking screws significantly decreased both varus/valgus angulation and medial/lateral translation under load, compared with the parallel screws. Constructs were 117% and 55% as rigid as the intact tibia in axial loading with and without cortical contact, and 6.5% and 3.1% as stiff in torsion.(ABSTRACT TRUNCATED AT 250 WORDS)