1. The discharge of single afferent chemoreceptor fibres was recorded from the cut sinus nerve over periods of 60 or 90 min of constant, isocapnic hypoxia (arterial O2 pressure, Pa,O2, 3.13-5.25 kPa), in twenty anaesthetized rabbits, after dividing the sympathetic supply to the carotid body. 2. Under control conditions, discharge after 60 min of hypoxia adapted to a mean (S.E.M.) of 71.95 (2.75)% of that attained at 5 min of hypoxia in twenty-three hypoxic experiments. This adaptation was more pronounced when Pa,O2 was lower than 4 kPa (30 Torr). 3. Domperidone (1 mg kg-1 bolus + 1 mg kg-1 h-1 infusion I.V.), increased normoxic afferent discharge by a mean of 142%. In ten experiments, discharge after 60 min of hypoxia adapted to a mean (S.E.M.) of 56.22 (+/- 3.40)% of that attained at 5 min of hypoxia which was significantly different from control hypoxic runs (P = 0.006). 4. In seven experiments propranolol (1 mg kg-1 bolus + 1 mg kg-1 h-1 infusion I.V.) did not affect the normoxic discharge. The mean adaptation of discharge after 60 min of hypoxia was to 77.43 (3.97)% of discharge attained at 5 min of hypoxia, which was not significantly different from control hypoxic runs (P = 0.34). 5. Under control conditions plasma [K+] increased steadily during 60 min of hypoxia, in fourteen experiments, from a mean of 2.76 (0.14) to 2.85 (0.12) mmol l-1 but this was not significant (P = 0.21). Domperidone (n = 6) did not affect plasma [K+] at any time, but after propranolol (n = 6) it increased by a mean (S.E.M.) of 0.39 (0.09) mmol l-1 (P = 0.01) in normoxia and by a further 0.62 (0.28) mmol l-1 (P = 0.08) at 60 min of hypoxia. 6. The results suggest that the adaptation of chemoreceptor discharge to hypoxia in the rabbit is not mediated by changes in plasma [K+]; in addition, endogenous dopamine, but not noradrenaline, contributes to the maintenance of chemoreceptor discharge in prolonged hypoxia.