A systematic approach to evaluate the contribution of individual residues occurring within the sequence Ser386-Pro-Phe-Arg-Ser-Val-Gln392 from bovine kininogen towards binding to tissue kallikrein is developed. Of the 21 sequences which can be formed, no dipeptide and only one tripeptide measurably inhibits the enzyme. Almost 80% of the binding energy of the substrate analogue inhibitors comes from the core sequence Phe-Arg-Ser which occurs between P2 and P1'. Molecular models developed from the Chen-Bode coordinates of the aprotinin--beta-PPK complex have been used to interpret the results of these studies.