Biomechanical testing was performed to evaluate five techniques of internal fixation of transverse patella fractures. Using cadaveric lower extremities, transverse osteotomies of the patella were performed, and the simulated fractures were fixed with the following techniques: the modified tension band, anterior tension band with a supplemental cerclage wire (the Pyrford technique), tension band with cancellous bone screws, Pyrford technique with cancellous screws, and cancellous screws alone. The fixation techniques were evaluated by measuring the separation of the fracture fragments during loading to produce a physiologic range of motion (90° flexion to full extension). All techniques functioned adequately, with no fracture gap exceeding 1 mm. The tension band with screws technique performed significantly better than did the modified tension band, with an average fracture gap approximately half that of the traditional modified tension band technique. Mechanically, the addition of the screws to the tension band techniques reduces fracture separation by providing compression throughout the range of motion and by resisting the tensile loading during terminal extension.