Proteins are large complex biomolecules that act as the effectors of essentially all cell functions. Due to the intrinsic complexity of protein architecture at the microscopic level and the inadequacy of theoretical methods to predict protein reactivity (ie, folding, stability, and function), protein engineering has emerged as a valuable tool to investigate structure–stability–activity relationships in proteins and nowadays recombinant DNA technologies are the “gold standard” for site‐specifically manipulating a given protein chain. The usefulness of current mutagenesis techniques, however, is limited by the relatively poor chemical diversity of the 20 DNA‐coded amino acids, such that it is difficult to precisely assign the observed change of protein stability or function to the variation of a single physicochemical property at a protein site (ie, hydrophobicity, conformational propensity, polarizability, hydrogen bonding, etc). In this article, we report relevant examples from our laboratory showing that chemical methods, that is, enzyme‐catalyzed semisynthesis and stepwise solid‐phase synthesis, allow to conveniently incorporate non‐natural amino acids with “tailored” side chains into small proteins and thus effectively transfer the structure–activity relationship methodology, typical of the medicinal chemistry approach on small molecules, to the study of folding, stability, and molecular recognition in macromolecular protein systems.