During the 1980s the purification, cloning, and expression of various forms of guanylyl cyclase (GC) revealed that they served as receptors for extracellular signals. Seven membrane forms, which presumably exist as homodimers, and four subunits of apparent heterodimers (commonly referred to as the soluble forms) are known, but in animals such as nematodes, much larger numbers of GCs are expressed. The number of transmembrane segments (none, one, or multiple) divide the GC family into three groups. Those with no or one transmembrane segment bind nitric oxide/carbon monoxide (NO/CO) or peptides. There are no known ligands for the multiple transmembrane segment class of GCs. Mutational and structural analyses support a model where catalysis requires a shared substrate binding site between the subunits, whether homomeric or heteromeric in nature. Because some cyclases or cyclase ligand genes lack specific GC inhibitors, disruption of either has been used to define the functions of individual cyclases, as well as to define human genetic disease counterparts.