Flavocytochrome b2, a flavohemoprotein, catalyzes the oxidation of lactate at the expense of the physiological acceptor cytochrome c in the yeast mitochondrial intermembrane space. The mechanism of electron transfer from the substrate to monoelectronic acceptors via FMN and heme b2 has been intensively studied over the years. Each prosthetic group is bound to a separate domain, N-terminal for the heme, C-terminal for the flavin. Each domain belongs to a distinct evolutionary family. In particular, the flavodehydrogenase domain is homologous to a number of well-characterized l-2-hydroxy acid-oxidizing enzymes. Among these, some are oxidases for which the oxidative half-reaction produces hydrogen peroxide at the expense of oxygen. For bacterial mandelate dehydrogenase and flavocytochrome b2, in contrast, the oxidative half-reaction requires monoelectronic acceptors. Several crystal structures indicate an identical fold and a highly conserved active site among family members. All these enzymes form anionic semiquinones and bind sulfite, properties generally associated with oxidases, whereas electron transferases are expected to form neutral semiquinones and to yield superoxide anion. Thus, flavocytochrome b2 is a highly unusual dehydrogenase-electron transferase, and one may wonder how its flavin reacts with oxygen. In this work, we show that the separately engineered flavodehydrogenase domain produces superoxide anion in its slow reaction with oxygen. This reaction apparently also takes place in the holoenzyme when oxygen is the sole electron acceptor, because the heme domain autoxidation is also slow; this is not unexpected, in view of the heme domain mobility relative to the tetrameric flavodehydrogenase core (Xia, Z. X., and Mathews, F. S. (1990) J. Mol. Biol. 212, 837-863). Nevertheless, this reaction is so slow that it cannot compete with the normal electron flow in the presence of monoelectronic acceptors, such as ferricyanide and cytochrome c. An inspection of the available structures of family members does not provide a rationale for the difference between the oxidases and the electron transferases.