The weak intermolecular forces inherent to molecular crystals give rise to polymorphism, or the ability for these materials to adopt multiple solid-state packing arrangements. Differences in molecular packing can result in distinct materials properties of the same compound (e.g., diamond vs. graphite), making polymorphism of general interest to the materials community. Yet, our collective understanding of this phenomenon remains inadequate. Calculations can shed light on the energetic landscape across different polymorphs but predictions for which phases are preferentially accessed remain challenging because accessibility is often dictated by kinetics (i.e., processing), as opposed to thermodynamics (i.e., energetics). In this talk, I will highlight a framework we have developed that directly correlates the presence of short intermolecular contacts with polymorphic accessibility. Subtle changes in the molecular chemistry do not significantly affect the possible solid-state packing arrangements, but do result in differences in the non-covalent intermolecular interactions that ultimately lead to preferential access to specific polymorphs. Starting with a series of core-chlorinated naphthalene diimide derivatives, we show this framework to be widely applicable to compounds ranging from molecular semiconductors to pharmaceutics, such as rotinavir, and biological building blocks, such as guanine. In this era of machine learning and materials by design, this framework can identify which polymorphs, among those available, are practically accessible and stable, thereby extending the predictive power of the solid-state properties of molecular materials prior to synthesis.