The instability of human myeloperoxidase [EC 1.11.1.7] compound I, which was spontaneously reduced to compound II, and the abnormal stoichiometry of the reaction of myeloperoxidase with H2O2 were investigated. As to the former, a pretreatment of myeloperoxidase with H2O2 did not stabilize compound I, and no difference in its stability was observed between native (alpha 2 beta 2) and hemi (alpha beta) myeloperoxidase. From these results, it was thought that the instability of compound I was caused by neither the presence of endogenous donors nor the intramolecular reduction of compound I to compound II by the other heme in the native enzyme molecule. As for the latter, true catalase activity of myeloperoxidase was demonstrated by monitoring O2 evolution after the injection of H2O2 into the enzyme solution. Myeloperoxidase compound I reacted with H2O2 and returned to the ferric state with concomitant evolution of an O2 molecule. Accordingly, the abnormal stoichiometry of the reaction with H2O2 and a part of the instability of compound I can probably be ascribed to this true catalase activity.