We apply general thermodynamics-based wave analysis methods to a gas-gun-driven plate impact experiment designed to derive strength information from tantalum at pressures of 10–25 GPa. The analysis provides estimates of the complete deformation paths in terms of the coupled evolution of mean stress, deviatoric stress, plastic strain, and plastic strain rate, yielding detailed information for direct comparison to strength models. This inverse analysis (deriving estimates of strength behavior directly from the measurements, with no strength model assumed) is compared to forward analysis (hydrodynamic simulations with specific strength models, in general adjusting parameters to optimally match the experiment). This comparison fulfills three goals. (1) To determine the parameter sensitivity and overall stability of the inverse analysis by analyzing simulated data as if it were experimental data. We find that, in reasonably favorable cases, precision to ∼10% is possible for the flow curve during loading and ∼30% for the shape of the curve during unloading. (2) To distinguish the ability of different strength models to account for the measurements. In particular we find that a new multiscale strength model seems to capture the rate-dependent release behavior very well but that it is difficult to capture the effects of a particular material’s microstructure and texture. (3) To bracket our understanding of the actual strength behavior in the experiment and enhance our confidence in both the forward and inverse calculations. The results show a peak deviatoric stress of ∼0.7–1.4 GPa occurring nearly at the point of peak plastic strain rate, followed by a complex evolution in which the material’s internal relaxation and strain-hardening properties interact with the rest of the loading wave, the post-shock plateau, and the unloading wave. The results show the importance of extreme precision in measurement t- - iming and equation-of-state calibrations, particularly at higher pressures.