Large strides have been achieved in nanoparticle self-assembly, using various strategies to achieve ordered, supracolloidal structures, ranging from dimers to chains and vesicles to 3-D lattices. However, these methods, while expanding the scope and accessibility of design, face inherent limitations in targeting complex structures with high yields, particularly when using isotropic building blocks (e.g. gold nanoparticles and polystyrene nanoparticles). Additionally, research studying the reversibility of nanoparticle assemblies is mostly limited to small-ligand-modified particles rather than polymer-modified nanoparticles. Polymers are particularly advantageous as they provide a higher degree of functionality to the nanoparticle surface and allow for increased control in directing particle interactions. This control is necessary to continue furthering the advancement of gold nanoparticles in plasmonics, sensors, and catalysts. Here, we introduce two strategies to assemble gold nanoparticles into supracolloidal nanostructures. Gold nanoparticles are modified with complementary, functionalized-block-copolymers that drive the assembly of the nanoparticles. The first strategy uses a diblock copolymer composed of a hydrophilic outer block and an acid or base-functionalized inner block. Upon mixing, particles are assembled due to the acid–base neutralization between the complementary block copolymers. The resultant supracolloids consist of nanoparticles precisely arranged in space, which mimic the geometries of small molecules. The particle interactions are fine tuned by varying the size and feeding ratio of the nanoparticles, along with the length and composition of the block copolymers. Careful tuning of these parameters yields nanostructures with different valences that were produced in high yield. Additionally, the implementation of a long outer, hydrophilic polymer block provided the assembled nanostructures stability when transferred from THF to water. Colloidal stability in an aqueous medium could allow for expanded use of these nanostructures in cellular uptake studies and biomedical applications. The second strategy uses a diblock copolymer composed of a hydrophilic outer block and an inner block containing either complementary host or guest moieties. Particularly, we take advantage of the well-established interactions between β-cyclodextrin and adamantane as the host and guest molecules. Upon the slow addition of water, particles assemble due to the host–guest interactions between the complementary block copolymers, as the hydrophobic adamantane moieties are driven within the β-cyclodextrin macrocycles. Fine tuning of the nanoparticle sizes and feeding ratios and the block copolymer lengths and compositions results in high yields of targeted supracolloids that also mimic the geometries of molecules. Interestingly, the size difference between the host and guest-modified particles led to different types of nanostructures. In addition, due to the reversibility of the host–guest interactions, we demonstrate the ability of our system to reorder in response to competitive host moieties. Upon addition of free β-cyclodextrin, the host–guest interactions are disrupted, resulting in disassembly of the nanostructures, which we could reassemble upon removal of the free cyclodextrin. Finally, due to the strength of the nanoparticle interactions, we also tested the selectivity of the nanoparticle interactions by assembling the host building block with different guest building blocks. We showed that when assembled with competing guest building blocks, the β-cyclodextrin building blocks preferentially interact the adamantane building blocks due to the stronger particle interactions. This reversibility and selectivity make our system a potential candidate for use in biosensors.