To some it may seem that we now know less about respiratory chemoreception than we did 20 years ago. Back then, it was widely accepted that the central respiratory chemoreceptors (CRCs) were located exclusively on or near the surface of the ventrolateral medulla (VLMS). Now, instead, it is generally believed that there are widespread sites of chemoreception, and there is little agreement on when and how each of these sites is involved in respiratory control. However, those in the field know that this actually is progress, primarily because we have gone from simply identifying candidate regions, to identifying specific neuronal subtypes that may be the sensors. In this invited review, we have been asked to discuss some of the current controversies in the field. First, we define the minimal requirements for a cell to be a CRC, and what assumptions can not be made without more data. Then we review the evidence that two neuronal subtypes, serotonergic neurones of the midline raphe and glutamatergic neurones of the retrotrapezoid nucleus, are chemoreceptors. There is evidence supporting a role in respiratory chemoreception for both types of neurone, as well as the other candidates, but there is also information that is missing. Future work will need to focus on which of the candidates are indeed chemoreceptors, what percentage of the overall response each one contributes, and how this percentage varies under different conditions.