This article develops the possibility of using conformationally rigid peptides as bioorganic model systems. Stereochemical constraints on peptide backbone folding may be introduced by the judicious use of sequences containingα-aminoisobutyric acid and proline. The design of synthetic peptide models of 3₁₀-helical and β-turn conformations is reviewed. Attempts at generating model antiparallelβ-sheet peptides are discussed. The use of disulphide crosslinks is illustrated in the application of cystine peptides to generate models forβ-turn and antiparallelβ-sheet conformations. Using a conformationally well-defined backbone as a skeleton attempts to generate models for protein binding sites are examined. Helical retinylidene-lysine peptides are introduced as models for the bacteriorhodopsin chromophore. Lysine containing peptides and chiral diamines are explored as model-binding sites for bilirubin and gossypol. An attempt to model the active site disulphide loop of the redox protein, thioredoxin, is described.