Realistic shock loading scenarios imply combined, multiaxial stress states, where the mechanically induced energy competes for failing the material intrinsically due to cohesive failures. Under shock loading, failure encompasses various modes, including spallation, shear bands, microcracking, morphological alteration, phase transformation, and plastic deformation. However, elucidating failure in high strain loading conditions challenges the state-of-the-art characterization due to the disparity interfacing different loading and measurement techniques. This research used a novel spectro-mechanical characterization setup, integrating bulk terahertz spectroscopy synchronized in real-time with mixed-mode laser-induced shock waves. The latter was used to submit polyurea elastomers to concurrent pressure and shear waves by saddling a thin polymer sheet on the hypotenuse of a glass prism. Hence, the laser-induced pressure shock waves underwent simultaneous mode conversion at the glass-polyurea interface at the hypotenuse of the glass prism, effectively submitting the sample to tunable amplitude pressure-shear shock waves. The terahertz spectroscopy, operating in the right-angle reflection configuration, was modified to probe predetermined spots on the polyurea samples at different stress states (adjusted by changing the illumination energy) in a three-step sequence: before, during, and after shock loading. The shock-loaded polyurea samples were also characterized using the optical and scanning electron microscopes, revealing compression-shear and tensile-shear failure modes, including surface depression, nucleation of micro-voids and microcracks, and formation of small planes. The induced stress state based on the illumination energy was quantified using an accompanying finite element simulation, resulting in completing the quasi-isentropic behavior of polyurea elastomers, which was found to be in good agreement with previous experimental data from the plate impact experiments and the predictions based on the Lennard-Jones potential. Time domain analysis of terahertz data, using temporal shifting, hysteresis, and dynamic time warping based on the sum of the Euclidean distances between signal pairs, affirmed the microscopy observations while shedding insights on the elastic and plastic deformations and the reversible conformational changes. Irreversible changes to the macromolecule were also detected using the terahertz data spectral analysis. The current results establish a new experimental framework for fundamentally characterizing polymers and their composites at ultrahigh strain rate loading conditions.