Caspases comprise a family of dimeric cysteine proteases that control apoptotic programmed cell death and are therefore critical in both organismal development and disease. Specific inhibition of individual caspases has been repeatedly attempted, but has not yet been attained. Caspase-9 is an upstream or initiator caspase that is regulated differently from all other caspases, as interaction with natural inhibitor XIAP-BIR3 occurs at the dimer interface maintaining caspase-9 in an inactive monomeric state. One route to caspase-9-specific inhibition is to mimic this interaction, which has been localized to the α5 helix of XIAP-BIR3. We have developed three types of stabilized peptides derived from the α5 helix, using incorporation of amino-isobutyric acid, the avian pancreatic polypeptide-scaffold or aliphatic staples. The stabilized peptides are helical in solution and achieve up to 32 µM inhibition, indicating that this allosteric site at the caspase-9 dimerization interface is regulatable with low-molecular weight synthetic ligands and is thus a druggable site. The most potent peptides against caspase-9 activity are the avian pancreatic polypeptide-scaffolded peptides. Other caspases, which are not regulated by dimerization, should not be inactivated by these peptides. Given that all of the peptides attain helical structures but cannot recapitulate the high-affinity inhibition of the intact BIR3 domain, it has become clear that interactions of caspase-9 with the BIR3 exosite are essential for high-affinity binding. These results explain why the full XIAP-BIR3 domain is required for maximal inhibition and suggest a path forward for achieving allosteric inhibition at the dimerization interface using peptides or small molecules.