Plasma amine oxidases catalyze the oxidative deamination of amines to aldehydes, followed by a 2e- reduction of O2 to H2O2. Pyrroloquinoline quinone (PQQ), previously believed to be restricted to prokaryotes, has recently been proposed to be the cofactor undergoing reduction in the first half-reaction of bovine plasma amine oxidase (Ameyama, M., Hayashi, U., Matsushita, K., Shinagawa, E., and Adachi, O. (1984) Agric. Biol. Chem. 48, 561-565; Lobenstein-Verbeek, C. L., Jongejan, J. A., Frank, J., and Duine, J. A. (1984) FEBS Lett. 170, 305-309). This result is unexpected, since model studies with PQQ implicate Schiff's base formation between a reactive carbonyl and substrates, whereas experiments with bovine plasma amine oxidase have failed to provide evidence for a carbonyl cofactor. We have, therefore, re-examined putative adducts between substrate and enzyme-bound cofactor, employing a combination of [14C]benzylamine and [3H]NaCNBH3. The use of the relatively weak reductant, NaCNBH3, affords Schiff's base specificity and permits the study of enzyme below pH 7.0. As we show, enzyme can only be inactivated by NaCNBH3 in the presence of substrate, leading to the incorporation of 1 mol of [14C]benzylamine/mol of enzyme subunit at complete inactivation. By contrast, we are unable to detect any labeling with [3H]NaCNBH3, analogous to an earlier study with [3H]NaCNBH4 (Suva, R. H., and Abeles, R. H. (1978) Biochemistry 17, 3538-3545). We conclude, first, that our inability to obtain adducts containing both carbon 14 and tritium rules out the reductive trapping either of amine substrate with pyridoxal phosphate or of aldehyde product with a lysyl side chain and, second, that the observed pattern of labeling is fully consistent with the presence of PQQ at the active site of bovine plasma amine oxidase.