A high hardness armour steel (HHA) has been subjected to mechanical characterization under tension, compression, and shear loading at quasi-static and dynamic rates incorporating ambient and elevated temperatures. The resulting data has been used to derive constants for four plasticity constitutive models: Johnson-Cook (JC), Zerilli-Armstrong (ZA), modified Johnson-Cook (MJC), and a generalized J2-J3 yield surface (GYS). The resulting models have been used to predict the response of the HHA material during Taylor rod-on-anvil experiments. High speed photography and digital image correlation was used during the rod-on-anvil experiments to capture both transient deformation profiles and maximum principal strain along the surface of the rod (i.e. compression along the length of the rod). The JC, MJC, and GYS models were found to provide the best prediction of the shape of the rod (nose diameter and length), within 2% of the experimental measurement in all four rod-on-anvil experiments which did not result in fracture. The JC and GYS models, furthermore, were found to provide the best agreement with the measured transient surface strain profiles, predicting the experimental measurement to within 10% at all measurement locations and time steps for the experiment resulting in maximum deformation (impact velocity = 208 m/s). The results suggest that the added complexity of models such as the MJC and GYS, which incorporate strain hardening saturation, two-part strain rate dependency, and J3 plasticity effects, are unnecessary for HHA under the loading conditions experienced during rod-on-anvil experiments.