The influence of halothane 0.8-1.2% inspired on the peripheral hypoxic chemoreflex was investigated in 13 cats subjected to artificial brain stem perfusion (ABP). This technique allows for an independent control of blood gas tensions and halothane concentration between blood perfusing the brain stem (central) and the systemic circulation (peripheral). In six cats the ventilatory response to isocapnic hypoxia was assessed during overall halothane anesthesia (HO) before and during ABP. Before ABP, systemic and brain stem circulations both were rendered hypoxic. During ABP, hypoxia was induced systemically while the brain stem was maintained hyperoxic. The ventilatory response in non-ABP cats (mean 698 ml . min-1 at PaO2 6.6 kPa; 50 mmHg) was about half the response in ABP cats (mean 1,194 ml . min-1 at PaO2 6.5 kPa; 49 mmHg), indicating that in the presence of halothane, central hypoxia depressed ventilation appreciably. Compared with chloralose-urethane anesthesia (CU), halothane reduced the ventilatory response to hypoxia in both perfusion conditions but never abolished it. To assess the influence of halothane on peripheral and central mediation of the CO2 response during hypoxia, each was assessed during CU anesthesia, during HO, and with halothane applied exclusively peripherally against a background of CU (CUHP). In all drug states, the periphery was kept hypoxic and brain stem hyperoxic. Compared with CU anesthesia, HO and CUHP anesthesia reduced both peripheral (Sp) and central (Sc) CO2 sensitivity but not the Sp/Sc ratio. Similarly, the extrapolated PaCO2 at zero ventilation was not detectably different among these three states.(ABSTRACT TRUNCATED AT 250 WORDS)