The ferric reductase superfamily comprises several oxidoreductases that use an intracellular electron source to reduce an extracellular acceptor substrate. NADPH oxidases (NOXs) and six-transmembrane epithelial antigen of the prostate enzymes (STEAPs) are iconic members of the superfamily. NOXs produce extracellular reactive oxygen species that exert potent bactericidal activities and trigger redox-signaling cascades that regulate cell division and differentiation. STEAPs catalyze the reduction of extracellular iron and copper which is necessary for the bioavailability of these essential elements. Both NOXs and STEAPs are present as multiple isozymes with distinct regulatory properties and physiological roles. Despite the important roles of NOXs and STEAPs in human physiology and despite their wide involvement in diseases like cancer, their mode of action at the molecular level remained incompletely understood for a long time, in part due to the absence of high-resolution models of the complete enzymes. Our two laboratories have elucidated the three-dimensional structures of NOXs and STEAPs, providing key insight into their mechanisms and evolution. The enzymes share a conserved transmembrane helical domain with an eye-catching hourglass shape. On the extracellular side, a heme prosthetic group is at the bottom of a pocket where the substrate (O2 in NOX, chelated iron or copper in STEAP) is reduced. On the intracellular side, the inner heme of NOX and the FAD of STEAP are bound to topological equivalent sites. This is a rare case where critical amino acid substitutions and local conformational changes enable a cofactor (heme vs FAD) swap between two structurally and functionally conserved scaffolds. The catalytic core of these enzymes is completed by distinct cytosolic NADPH-binding domains that are topologically unrelated (a ferredoxin reductase-like flavoprotein domain in NOX and a F420H2:NADP+-like domain in STEAP), feature different quaternary structures, and underlie specific regulatory mechanisms. Despite their differences, these domains all establish electron-transfer chains that direct the electrons from NADPH to the transmembrane domain. The multistep nature of the process and the chemical nature of the products pose considerable problems in the enzymatic assays. We learned that great care must be exerted in the validation of a candidate inhibitor. Multiple orthogonal assays are required to rule out off-target effects such as ROS-scavenging activities or nonspecific interference with the enzyme redox chain. The structural analysis of STEAP/NOX enzymes led us to further notice that their transmembrane heme-binding topology is shared by other enzymes. We found that the core domain of the cytochrome b subunits of the mitochondrial complex III and photosynthetic cytochrome b6f are closely related to NOXs and STEAPs and likely arose from the same ancestor protein. This observation expands the substrate portfolio of the superfamily since cytochromes b act on ubiquinone. The rigidly packed helices of the NOX/STEAP/cytochrome b domain contrast with the more malleable membrane proteins like ion channels or amino-acid transporters, which undergo large conformational changes to allow passage of relatively large metabolites. This notion of a rigid hourglass scaffold found an unexpected confirmation in the observation, revealed by structural comparisons, that an helical bundle identical to the NOX/STEAP/cytochrome b enzymes is featured by a de novo designed heme-binding protein, PS1. Apparently, nature and protein designers have independently converged to this fold as a versatile scaffold for heme-mediated reactions. The challenge is now to uncover the molecular mechanisms that implement the isozyme-specific regulation of the enzyme functions and develop much needed inhibitors and modulators for chemical biology and drug design studies.