The precise control of breathing is fundamental to the survival of all terrestrial vertebrates. Breathing fulfils two essential functions: provision of oxygen to support metabolism; and removal of CO2, the unavoidable by-product of metabolism. CO2 readily combines with H2O to form HCO3- and H+. Because an increase by only 100 nM of the hydrogen ion concentration in blood can prove fatal, the regulated excretion of CO2 via breathing is extremely important for life. The accepted explanation for the nature of this regulation involves the indirect effect of CO2 on blood pH being detected by chemosensors in the brain stem that are highly sensitive to pH. Our work challenges this explanation and suggests that molecular CO2 can be directly detected.We discovered that ATP release in the central chemosensitive areas of the brain stem was a key signalling step in response to elevated CO2 (1). By following the mechanisms of ATP release we documented that hemichannels of connexin26 (Cx26) were an important conduit for CO2-dependent ATP release (2, 3). Furthermore our evidence suggested that CO2 interacted directly with Cx26 to open the hemichannel and allow the efflux of ATP (3). The ATP, once released, can excite specific neurons in the brain stem through activation of receptors selectively sensitive to ATP (4, 5). ATP release is thus the essential link between the opening of Cx26 by CO2, neuronal activation, and the consequent neurophysiological responses that result in enhanced breathing. Most recently we have provided compelling genetic evidence for the molecular mechanism of direct sensing of CO2 via Cx26 (6). The three known CO2-sensitive connexins (Cx26, Cx30, Cx32) possess a “carbamylation motif” that is absent from a CO2-insensitive (Cx31) connexin. This motif comprises Lys125 and four further amino acids that orient Lys125 towards Arg104 of the adjacent subunit of the connexin hexamer. By introducing the carbamylation motif into Cx31, we created a mutant hemichannel (mCx31) that was opened by increases in CO2. Mutation of the carbamylation motif in Cx26 and mCx31 destroyed CO2 sensitivity. Course-grained computational modelling of Cx26 demonstrated that the proposed carbamate bridge between Lys125 and Arg104 biases the hemichannel to the open state. Carbamylation of Cx26 introduces a new transduction principle for physiological sensing of CO2. In our latest studies, my laboratory has used conditional transgenesis to delete Cx26 from specific lineages of cells in intact mice, such as glial or neural crest-derived cells. These deletions greatly reduce CO2-dependent ATP release from the chemosensitive areas of the brain stem. Whole body plethysmography recordings from awake mice show that their sensitivity to variations of inspired CO2 is also greatly reduced. Our studies have helped to establish direct sensing of CO2 via connexin hemichannels as a new chemosensory transduction principle in the field in mammalian physiology.