Stable trochanteric fractures were produced in 12 pairs of human cadaver femora, which were fixed by either a 135 degree Jewett nail plate, a 135 degree NoLok sliding screw/plate, or a 140 degree Hansson pin/plate. The bone-implant preparations were then subjected to 20,000 load cycles simulating full weight bearing, measuring the elastic and permanent fracture displacement. In each case, elastic displacement occurred during loading. For the Jewett-stabilized fractures, a steady increased elastic displacement, as well as permanent displacement, was seen throughout the test, causing 2/8 failures when the implant penetrated the femoral head. In the NoLok- and Hansson-stabilized fractures, there were no failures, both giving better overall stability than the Jewett device with sufficient fracture stability throughout the test.