Glycosylation is a complex and ubiquitous post-translational modification that critically contributes to the structural and functional diversity of proteins and lipids. It occurs when glycans are enzymatically linked to specific amino acid residues on proteins, influencing folding, structural stability and function. In glycoconjugates, sialylation has been often shown to be an essential feature for molecular recognition and immune regulation, meanwhile altered sialylation patterns have been linked to cancer, infectious diseases, and neurological and inflammatory disorders. The molecular factors that determine whether sialylation has activating or inhibitory effects remain poorly defined. This thesis addresses this knowledge gap by exploring at the atomistic level of detail the impact of sialylation in a variety of molecular systems linked to different biological pathways and responses, key to human health and disease. To do this I used molecular dynamics (MD) simulations, a powerful computational approach that allows users to determine the structure, dynamics and energetics features underpinning molecular recognition and function through sampling and statistical thermodynamics principles. The results of the MD simulations I present are used to guide and/or to support experimental methods, leading to a common, clear rationale of the biological event examined. Through the studies I present in this thesis, I demonstrate that the effects of sialylation depend on its position, whether on receptors, peptide ligands, or glycolipids, and on its mode of presentation, such as monovalent or multivalent, and in soluble or in membrane-anchored constructs. I show that sialic acids can act as local inhibitors, for example by introducing steric hindrance that reshapes ligand preference at the endocytic receptor LRP1, which is relevant to neurodegenerative disease. Conversely, hypersialylation can function as a structural switch that promotes and stabilises high-avidity interactions, as observed in the CD52–HMGB1–Siglec-10 immune signalling pathway. Recognition by immunomodulatory receptors, such as Siglec-6, depends on the combined influence of the sialylated epitope and its orientation with respect to the lipid membrane. Sialylation of N-glycans is ultimately controlled by upstream modifications, such as the bisecting GlcNAc motif, which, as a stop codon, can halt further glycan maturation. By providing an atomistic view, the work in this thesis reveals that sialylation behaves in complex and context-dependent ways. When factors such as local environment, valency and the 3D shape of the sialylated epitope are considered, sialic acids emerge as active determinants of recognition rather than passive terminal caps. These results link the crucial role of sialylation to disease-relevant mechanisms and offer a structural framework to guide future glycan-focused therapeutic strategies.