Lantibiotics are (methyl)lanthionine-containing bacterial peptides. (Methyl)lanthionines are posttranslationally introduced into the prepropeptides by biosynthetic enzymes that dehydrate serines and threonines and couple these dehydrated residues to cysteine residues. Thirty seven lantibiotic primary structures have been proposed to date, but little is known about the substrate specificity of the lantibiotic modifying enzymes. To define rules for the rational design of modified peptides, we compared all known lantibiotic structures by in silico analysis. Although no strict sequence motifs can be defined that govern the modification, statistical analysis demonstrates that dehydratable serines and threonines are more often flanked by hydrophobic than by hydrophilic amino acids. Serine residues escape dehydration more often than threonines. With these rules, novel hexapeptides were designed that either were predicted to become modified or will escape modification. The hexapeptides were fused to the nisin leader and expressed in a Lactococcus lactis strain containing the nisin modifying and export enzymes. The excreted peptides were analyzed by mass spectrometry. All designed fusion peptides were produced, and the presence or absence of modifications was found to be in full agreement with the predictions based on the statistical analysis. These findings demonstrate the feasibility of the rational design of a wide range of novel peptides with dehydrated amino acid residues.