The participation of a serine endopeptidase, previously shown to be involved in endogenous cholecystokinin inactivation [Rose, Camus and Schwartz (1989) Neuroscience 29, 583-594], in the hydrolysis of various exogenous cholecystokinin peptides was studied with slices from rat cerebral cortex. In order to protect intermediate fragments from further degradation and mimick experimental conditions in this previous study, most experiments were performed in the presence of Thiorphan, an enkephalinase inhibitor, and bestatin, an aminopeptidase inhibitor, which did not significantly affect the rate of cholecystokinin-8 hydrolysis. All peptide fragments formed after incubation of cholecystokinin-8, non-sulphated cholecystokinin-8, cholecystokinin-6, cholecystokinin-5, cholecystokinin-4 or Asp-Tyr-Met-Gly-Trp were identified by isocratic high-performance liquid chromatography in several systems, fluorescence spectra and/or amino acid analysis. When identified, the appearing fragments were quantified by u.v. spectrophotometry and found to fully account for the substrate disappearance. The hydrolysis rate was higher for short cholecystokinin peptides than for the octapeptide and was, in all cases, diminished by 30-50% in the presence of diisopropyl fluorophosphate, a serine peptidase inhibitor. One of the main hydrolysis products of cholecystokinin-8, or its non-sulphated analogue, was cholecystokinin-5, whose formation was impaired in the presence of diisopropyl fluorophosphate. Cholecystokinin-5 itself was apparently a substrate for a serine peptidase leading to the formation of the tripeptide Gly-Trp-Met, later cleaved into Trp-Met and Trp. Hence a serine endopeptidase(s) appears to be responsible for cleavage of the two peptides bonds of the cholecystokinin-8 molecule where the carboxyl group is donated by a methionine residue.2+n addition,