Several non-peptide systems have been designed to mimic different types of reverse turns. The incorporation of some of these mimetics into biologically active peptides has led to peptidomimetics with enhanced activity or metabolic stability. This paper reports the conformational analysis of tetrapeptides containing several bicyclic mimetics, sequences containing proline, other N-methyl and N-hydroxy amino acids, and pipecolic acid at residue i + 2 of the turn, and control peptide sequences using the Monte Carlo/stochastic dynamics simulation with the new set of AMBER* parameters for proline-containing peptides in water as implicitly represented by the GB/SA solvation model. Simple N-methylation (Pro-D-NMeAA and D-Pro- NMeAA) and N-hydroxylation of the amide bond between residues i + 1 and i + 2 or inclusion of the larger ring homolog pipecolic acid (D-Pro-Pip) in the third position (i + 2) causes significant nucleation of reverse-turn structures. Spirotricycle analogs restrict three of the four torsion angles that characterize the type II β-turn. Spirolactum analogs also restrict two of the four torsion angles as effective β-turn constraints. However, the geometry of a turn induced by indolizidinone and BTD differs significantly from that of an ideal β-turn and (S)-indolizidinone is more effective as a reverse turn than as a β-turn mimetic. These systems provide useful conformational constraints when incorporated into the structure of selected bioactive peptides. Such analogs can scan receptors for biological recognition of β-turn scaffolds with oriented side chains through combinatorial libraries to efficiently develop three-dimensional structure - activity relationships.