Mesoscale self-assembly is a central route to creating novel metamaterials for a wide variety of applications. However, the polydispersity of nanoparticles or colloidal building blocks often adds an additional layer of complexity to the assembly behavior of mesoscopic sys- tems. This complexity can inhibit the straightforward implementation of materials design and significantly influences both the crystallization process and the stability of the resultant structure. In-depth comprehension of these effects is necessary for a fundamental under- standing of structure formation in mesoscopic systems. The impact of size dispersity on the final assembled structures remains poorly explored and understood, in particular for struc- tures other than the most common dense sphere packing structure types. To address this, we utilize molecular dynamics simulations to model the self-assembly of two types of parti- cles with different sizes, which interact through otherwise equivalent isotropic, double-well pair potentials. We examine the level of crystallinity to distinguish between ordered and disordered structures, as well as differentiate between distinct crystal structures as the size dispersity of the simulated systems increases. We find that the demixing of particles with different sizes occurs at similar size ratios across crystal structure types. Moreover, we also note that each structure’s sensitivity to the size ratio with respect to their degree of order is different, with some structures presenting more robustness than others. This work suggests that a profound understanding and strategic manipulation of size dispersity could serve as a powerful tool in governing the self-assembly of soft materials, leading to the design of more robust structures with tailored properties.