Peptides represent a promising format for the development of therapeutics since they combine the advantages of proteins and small molecules. Several techniques based on rational design or in vitro evolution can be used to develop peptides as therapeutics. A limitation of peptides is a rapid clearance from the blood circulation when applied intravenously, preventing their application in therapies that require prolonged drug exposure. Several methods have been tested to extend the in vivo half-life of peptides, including fusion to long-lived serum proteins, PEGylation and conjugation to non-covalent albumin-binding ligands. A particularly attractive approach is a “piggy back” strategy in which peptides bind via a ligand to albumin without increasing significantly their small size. Acylation with fatty acids is the most successful strategy for delaying peptide clearance. The attachment of either myristic or palmitic acid to insulin and GLP-1 led to the approval of three acylated peptide drugs as daily injectable. Nevertheless, acylated conjugates generally present low solubility and weak affinity for albumin. Peptides have been explored as an alternative, as their format offers multiple approaches for affinity maturation as well as an easier synthesis of the conjugates. However, the albumin-binding peptides developed so far do not overcome the disadvantages faced with fatty acids. The first aim of my project was the development of an albumin-binding tag that combines the favorable properties of both fatty acids and peptides, and overcomes at the same time their limitations. This tag was engineered in order to have i) a high affinity for human albumin, ii) a high solubility in physiological solutions, and iii) an efficient synthesis in conjunction with therapeutic peptides. Towards this goal, I created a chimeric format consisting of a fatty acid combined with a short linear peptide. After iterative rounds of affinity screening, an evolved peptide-fatty acid tag able to bind human, rat and rabbit albumin with Kd values of 39, 220 and 320 nM, respectively, was obtained. This makes it the best albumin-binding tag among all of the peptide- or fatty acid-based ligands that have been developed. This tag can be easily synthesized in conjugation with peptides on standard automated synthesizers. Finally, it exhibits a high solubility due to four negatively charged amino acids in the peptide sequence, despite the presence of a 16-carbon palmitic hydrophobic tail. The second aim of my project was the application of this tag to three bioactive bicyclic peptides, previously developed in the laboratory, in order to improve their in vivo half-lives. I demonstrated that the conjugation has, overall, a mild impact on peptides’ binding affinity towards their targets and human albumin, and increases their proteolytic stability. Pharmacokinetic experiments showed that the tag extends the elimination half-life of bicyclic peptides in rats and rabbits approximately 25-fold to over seven and five hours, respectively. Tag conjugation to a bicyclic peptide targeting coagulation factor XII led to an efficient inhibition of coagulation in rabbits up to eight hours, suggesting a potential application in humans as anti-thrombotic drug. These results indicated that the tag will efficiently prolong the exposure of target proteins to bioactive peptides in humans due to its higher affinity for human albumin than that of rat and rabbit.