Advances in the development of rolled-sheet magnetostrictive materials led to testing of a prototype wireless magneto-elastic torque (WiMET) sensor using the iron alloy Galfenol. As torque was applied to a shaft, stress-induced changes in the magnetic state of Galfenol that was bonded to the shaft were proportional to the applied torque. Building on that work, this thesis investigates strategies to improve both repeatability and the signal to noise ratio of WiMET sensor output. Multi-physics models of WiMET stress and magnetic states under applied torques are used to improve understanding of sensor operation. Testing to validate simulations is performed using Galfenol and Alfenol, a newer rolled-sheet alloy, for torsional loads of 0 – 200 in-lb, and under quasi-static and dynamic (0 – 2000 RPM) loading conditions. The experimental results presented support the potential of WiMET sensor use for dynamic torque measurement and health monitoring of drive train systems.