The next major step for the exoplanet community lies in the characterization of terrestrial exoplanets, especially when it comes to identifying biosignatures and determining the habitability of these worlds. In response to this, NASA has proposed the Habitable Worlds Observatory (HWO), with the primary goal of searching for and characterizing Earth-sized planets in the habitable zones (HZ) of their stars. However, current characterization strategies that only rely on the unpolarized flux from the planets lose some of the informational content of the observed light and therefore suffer from degeneracies in the calculated planetary parameters. The sensitivity of polarization to the micro- and macro-physical properties of a planet’s atmosphere and surface makes it an invaluable tool that will enhance the characterizations of these worlds. Having accurate theoretical models of terrestrial exoplanets, validated against observations of solar system planets with known properties, will be vital for assisting in these characterizations. In this dissertation, I explore the power of polarimetry in characterizing terrestrial exoplanets and demonstrate its effectiveness at distinguishing between habitable and non-habitable planetary scenarios. I performed the first benchmarking of two independent polarization-enabled radiative transfer (RT) codes, as well as conducted the first investigations of the changes in the Earth’s spectropolarimetric signatures across all four geologic eons. Additionally, as terrestrial exoplanets are expected to be heterogeneous like the Earth, I analyzed the effects of time-variable features such as diurnal rotation, seasonal weather patterns, and different surface configurations on the resulting spectropolarimetric signatures of Earth and early Mars. With these models, I provided preliminary observing constraints for differentiating the habitable and non-habitable planetary scenarios with the HWO. Finally, as advanced telescopes will soon necessitate distinguishing habitable and non-habitable HZ planets, I used RT models to assist with characterizing new ground based observations of the polarized reflected light from Earth’s non-habitable twin Venus.