Shape memory polymers have recently become the focus of research for their unique ability to switch between two modulus states, allowing them to both recover from large amounts of strain as well as support complex loads. Part of this research involves engineering new formulas specifically designed for applications where traditional thermally activated SMPs are not ideal by tailoring the activation method used to transition the polymer. One such class of polymers is those that utilize optical energy at specific wavelengths to create and cleave crosslinks. It is the development of this new class of light activated shape memory polymers (LASMP) that is the focus of the presented work.;Experimental methods are newly created for this novel class of active materials. Several candidate LASMP formulas are then subjected to this set of experiments characterizing their mechanical and optical properties. Experimentally observed variations among the formulae include virgin state modulus, percent change in modulus with stimulus, and in some instances inelastic response.;To expedite the development of LASMP, a first principles multi-scale model based on the polymer's molecular structure is presented and used to predict the stress response of the candidate formulas. Rotational isomeric state (RIS) theory is used to build a molecular model of a phantom polymer chain. Assessment of the resulting conformation is then made via the Johnson family of statistical distributions and Boltzmann statistical thermodynamics. The ability of the presented model to predict material properties based on the molecular structure of the polymer reduces the time and resources required to test new candidate formulas of LASMP as well as aiding in the ability to tailor the polymer to specific application requirements.;While the first principles model works well to identify promising formulas, it lacks precision. The stress contribution from the constraints on the polymer chain's junctions and neighboring chain entanglements is then added to that of the phantom network allowing Young's modulus to be calculated from the predicted stress response of the polymer. Simple extension, equi-biaxial, and shear strain states are modeled and associated predicted material properties presented. The added precision of this phenomenological extension will aid device design.