Two kinds of dipeptidyl aminopeptidase I (DAP I [cathepsin C])-like activities which hydrolyze Gly-Phe-p-nitroanilide (Gly-Phe-pNA) were detected in Pseudomonas sp. strain WO24. They were purified and characterized. The isolated enzymes, named DAP BII and DAP BIII, were revealed to be homogeneous by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and isoelectric focusing. DAP BII was estimated to have a molecular mass of 150,000 Da by gel filtration and a subunit size of 73,000 Da by SDS-PAGE, indicating it to be a homodimer. The molecular mass of DAP BIII was evaluated to be approximately 60,000 Da by gel filtration and 69,000 Da by SDS-PAGE, indicating that it is monomeric. The isoelectric points of DAP BII and DAP BIII were 6.1 and 5.0, and their optimal pHs were 8.0 and 8.5 to 9.0, respectively. The result of peptide mapping for DAP BII and DAP BIII showed that these enzymes consist of different components. Both enzymes were completely inhibited by diisopropylphosphofluoride but not by general thiol inhibitors, indicating that they are serine proteases. DAP BII and DAP BIII hydrolyzed Gly-Phe-pNA but not Gly-Arg-pNA, both of which are model substrates for mammalian DAP I. Despite these shared activities toward DAP I, DAP BII released dipeptides from Ala-Ala-pNA and Lys-Ala-4-methylcoumarinamide (a substrate for DAP II), whereas DAP BIII did not hydrolyze either of these compounds and was presumed to prefer substrates composed of bulky, hydrophobic amino acids at P1 and P1' positions. In addition, DAP BII showed no endopeptidase activity, whereas DAP BIII possessed the activity on N-terminally blocked peptide derivatives besides exopeptidase activity. Assays performed with bioactive peptides such as angiotensin I and neuromedin N as substrates indicate that DAP BII has a considerably broader substrate specificity than DAP BIII and is able to hydrolyze an X-Pro bond, an imido bond that few peptidases and no known DAPs can cleave. These characteristics, namely, substrate specificities, molecular mass, pI, peptide mapping, pH optimum, and effect of inhibitors, suggested that the two DAPs purified in this work are distinct enzymes and do not belong to any of the previously reported DAP classes.