I will briefly present ten pieces of evidence which supports the following conclusions: a) The retrotrapezoid nucleus (RTN) is an anatomically well-defined pH-modulated nucleus located at the ventral surface of the medulla oblongata and the probable cellular basis of the long-postulated ventral medullary surface central chemoreceptors. b) RTN stabilizes arterial PCO2/pH by regulating lung ventilation. c) RTN is especially important when breathing is in autorhythmic mode (anesthesia, quiet waking, non-REM sleep). d) RTN’s synaptic inputs are many and may be just as important for breathing regulation as its pH-sensitivity. The evidence: 1) All RTN neurons (~ 800 in mice, ~2000 in rats) express Vglut2, Phox2b and neuromedin B; 50% express galanin. RTN neurons innervate selectively the lower brainstem regions that contain the respiratory pattern generator. 2) Many RTN neurons express cFos in rodents exposed to hypercapnia but not hypoxia. 3) In anesthetized rats, RTN unit activity increases by 0.5 Hz for every 0.01 change in arterial pH; this response persists after pharmacological blockade of the respiratory pattern generator. 4) Optogenetic activation of RTN in conscious rodents increases inspiratory frequency and amplitude, elicits active expiration and controls post-inspiratory air flow. If Vglut2 is deleted from RTN neurons, their optogenetic activation no longer stimulates breathing. 5) Optogenetic inhibition of RTN in unanesthetized rats reduces breathing in direct proportion to arterial pH; notably, RTN inhibition has virtually no effect when arterial pH exceeds 7.5. 6) Optogenetic inhibition of RTN produces a massive reduction of breathing frequency and amplitude during quiet waking or non-REM sleep in rats but has no effect on breathing frequency during REM sleep. 7) In vitro, RTN is activated by [H+] (slices) and the effect persists after cell isolation. 8) The pH response in vitro requires expression of a proton-activated GPCR (GPR4) and a proton-modulated potassium channel (TASK-2). These proteins are undetectable in astrocytes and the rest of the lower brainstem respiratory network. 9) The central respiratory chemoreflex is greatly reduced in mice with genetic deletion of RTN neurons (via Phox2b mutations), and in GPR4 or TASK-2 global knock-out mice. Reintroducing GPR4 specifically into RTN neurons restores a normal chemoreflex in GPR4 KO mice. 10) RTN neurons receive powerful excitatory inputs from the hypothalamus and the spinal cord and inhibitory feedback form lung stretch receptors and the respiratory pattern generator.