We have developed a strategy for the synthesis of positional-scanning synthetic combinatorial libraries (PS-SCL) that does not depend on the identity of the P1 substituent. To demonstrate the strategy, we synthesized a tetrapeptide positional library in which the P1 amino acid is held constant as a lysine and the P4-P3-P2 positions are positionally randomized. The 6,859 members of the library were synthesized on solid support with an alkane sulfonamide linker, and then displaced from the solid support by condensation with a fluorogenic 7-amino-4-methylcoumarin-derivatized lysine. This library was used to determine the extended substrate specificities of two trypsin-like enzymes, plasmin and thrombin, which are involved in the blood coagulation pathway. The optimal P4 to P2 substrate specificity for plasmin was P4-Lys/Nle (norleucine)/Val/Ile/Phe, P3-Xaa, and P2-Tyr/Phe/Trp. This cleavage sequence has recently been identified in some of plasmin's physiological substrates. The optimal P4 to P2 extended substrate sequence determined for thrombin was P4-Nle/Leu/Ile/Phe/Val, P3-Xaa, and P2-Pro, a sequence found in many of the physiological substrates of thrombin. Single-substrate kinetic analysis of plasmin and thrombin was used to validate the substrate preferences resulting from the PS-SCL. By three-dimensional structural modeling of the substrates into the active sites of plasmin and thrombin, we identified potential determinants of the defined substrate specificity. This method is amenable to the incorporation of diverse substituents at the P1 position for exploring molecular recognition elements in proteolytic enzymes.