The Complex II family of enzymes, comprising the respiratory succinate dehydrogenases and fumarate reductases, catalyze reversible interconversion of succinate and fumarate. In contrast to the covalent flavin adenine dinucleotide (FAD) cofactor assembled in these enzymes, the soluble fumarate reductases (e.g. that from Shewanella frigidimarina) that assemble a noncovalent FAD cannot catalyze succinate oxidation but retain the ability to reduce fumarate. In this study, an SdhA-H45A variant that eliminates the site of the 8α-N3-histidyl covalent linkage between the protein and the FAD was examined. The variants SdhA-R286A/K/Y and -H242A/Y, that target residues thought to be important for substrate binding and catalysis were also studied. The variants SdhA-H45A and -R286A/K/Y resulted in assembly of a noncovalent FAD cofactor, which led to a significant decrease (−87 mV or more) in its reduction potential. The variant enzymes were studied by electron paramagnetic resonance spectroscopy following stand-alone reduction and potentiometric titrations. The “free” and “occupied” states of the active site were linked to the reduced and oxidized states of the FAD, respectively. Our data allows for a proposed model of succinate oxidation that is consistent with tunnel diode effects observed in the succinate dehydrogenase enzyme and a preference for fumarate reduction catalysis in fumarate reductase homologues that assemble a noncovalent FAD.