The known solid-state structure (Fig. 1, top) of cyclo(b-HAla) 4 was used to model the structure of the title compound 1 as a prospective somatostatin mimic (Fig. 1, bottom). The synthesis started with the N-protected natural amino acids Boc-Phe-OH, Boc-Trp-OH, Boc-Lys(2-Cl-Z)-OH, and Boc-Thr(OBn)-OH, which were homologated to the corresponding b-amino-acid derivatives (Scheme 1) and coupled to the b-tetrapeptide Bocb-HTrp-b-HPhe-b-HThr(OBn)-b-HLys(2-Cl-Z)-OMe (16); the (N-Me)-b-HThr-(N-Me)-b-HPhe analog 17 was also prepared. C-and N-terminal deprotection and cyclization through the pentafluorophenyl ester gave the insoluble b-tetrapeptide with protected Thr and Lys side chains ( 18). Solubilization and debenzylation could only be effected in LiCl-containing THF (ca. 10% yield; with ca. 55% recovery). HPLC Purification provided a sample of the title compound 1, the structure of which, as determined by NMR-spectroscopy (Fig. 2, left) was drastically different from the theoretical model (Fig. 1). There is a transannular H-bond dividing the macrocyclic 16-membered ring, thus forming a ten-and a twelve-membered H-bonded ring, the former mimicking, or actually being superimposable on, an a-peptidic so-called b-turn. Still, the four side chains occupy equatorial positions on the ring, as planned, albeit with somewhat different geometry as compared to the original. The cyclo-b-tetrapeptide has micromolar affinities to the human somatostatin receptors (hsst 1 ± 5). Thus, we have demonstrated for the first time that it is possible to mimic a natural peptide hormone with a small b-peptide. Furthermore, we have discovered a simple way to construct the ubiquitous b-turn motif with bpeptides (which are known to be stable to mammalian peptidases).