The recently discovered prokaryotic signal transducer HemAT, which has been described in both Archaea and Bacteria, mediates aerotactic responses. The N-terminal regions of HemAT from the archaeon Halobacterium salinarum (HemAT-Hs) and from the Gram-positive bacterium Bacillus subtilis (HemAT-Bs) contain a myoglobin-like motif, display characteristic heme-protein absorption spectra, and bind oxygen reversibly. Recombinant HemAT-Hs and HemAT-Bs shorter than 195 and 176 residues, respectively, do not bind heme effectively. Sequence homology comparisons and three-dimensional modeling predict that His-123 is the proximal heme-binding residue in HemAT from both species. The work described here used site-specific mutagenesis and spectroscopy to confirm this prediction, thereby providing direct evidence for a functional domain of prokaryotic signal transducers that bind heme in a globin fold. We postulate that this domain is part of a globin-coupled sensor (GCS) motif that exists as a two-domain transducer having no similarity to the PER-ARNT-SIM (PAS)-domain superfamily transducers. Using the GCS motif, we have identified several two-domain sensors in a variety of prokaryotes. We have cloned, expressed, and purified two potential globin-coupled sensors and performed spectral analysis on them. Both bind heme and show myoglobin-like spectra. This observation suggests that the general function of GCS-type transducers is to bind diatomic oxygen and perhaps other gaseous ligands, and to transmit a conformational signal through a linked signaling domain.