The in vitro response of the carotid body to low or zero glucose appears equivocal, being robust in co-cultures of Type I cells with petrosal neurons or in carotid body thin slices (1, 2) but absent in whole carotid body preparations (3). As this difference may be due to the metabolic status of the various preparations or to an O2-dependence, we measured chemoafferent responses to zero glucose at varying levels of PO2 and following prior exposure to either hypoxia or hyperoxia. Pairs of carotid bodies were isolated from adult (> 30 days), male Wistar rats during isoflurane inhalation anaesthesia (2.5% in O2) and single fibre recordings of chemoafferents were made from the sinus nerve as described previously (4). Whilst recordings were made from the first of each pair of carotid bodies the second organ was incubated in a gassed (95% O2/5% CO2) ice-cold HCO3- -buffered solution before it also was recorded from. The time of incubation was between 4-5 hrs. Superfusate glucose was either 11mM or 0mM. Superfusate PO2 was adjusted to set basal discharge at an in vivo, normoxic value of 0.5-1.0 Hz. Data is expressed as means ± S.E.M and significance measured using t-tests (Statview, Abacus Concepts) and taken as P<0.05. On removal of glucose from the superfusate, chemodischarge in the first carotid body remained unchanged at a basal frequency of 0.68±0.14 Hz for 32.8±1.9 min, at which time discharge frequency increased significantly to 2.70±0.19 Hz (P < 0.01), before failing irreversibly. This time was significantly reduced to 15.4±2.8 min (P < 0.001) by either 2 or 3 prior exposures of the carotid body to severe hypoxia. In the second, hyperoxia-incubated, carotid body, the time taken to respond to zero glucose was also significantly reduced to 19.8±1.4 min (P < 0.001). All preparations responded to reducing PO2 with exponential increases in discharge both in the presence and absence of glucose. However, in the first carotid body of each pair, the PO2 response curves in the absence of glucose were significantly left shifted relative to control but were right shifted in the second of each pair, with the PO2 needed to achieve a discharge rate of 5Hz being decreased significantly by 8.39±0.68 mmHg (P < 0.05) in the first and increased by 13.3±3.89 mmHg (P < 0.05) in the second preparation. Our data confirms previous findings that show that the freshly-isolated, intact carotid body has a naturally poor acute sensitivity to reduced glucose and shows that this is not a PO2 dependent effect. We suggest, however, that the direct glucose sensitivity of in vitro chemoreceptor tissue may be increased by prior exposure to repeated bouts of natural hypoxic stimulation or by incubation in high PO2 media.