The chemical control of breathing is mediated through central chemoreceptors, which detect brain tissue P_CO2/pH (1), and peripheral chemoreceptors, which detect arterial P_CO2 in a P_O2-dependent manor (2). However, the mathematical nature of the interaction between these chemoreceptors is controversial. Using a novel rat preparation, we demonstrated recently that the magnitude of the phrenic response to a single step of specific carotid body hypoxia (60 Torr P_O2) was modulated by the level of brainstem P_CO2 (25 vs. 50 Torr P_CO2; 3). Based on these data, we hypothesized that the interaction between brainstem and carotid body chemoreceptors in the rat is hypo-additive. To test this hypothesis, we evaluated the effects of brainstem P_CO2 on a broader range of carotid body activation, spanning from 400 to 40 Torr P_O2. We used an in situ, decerebrate, vagotomized, arterially perfused rodent preparation (male Sprague-Dawley albino, decerebrated following halothane inhalation overdose). In this preparation, central and peripheral chemoreceptors are independently perfused with defined medium containing precisely controlled gas mixtures (4) and phrenic nerve activity is used to assess ventilatory responses. Five minute isocapnic (35 Torr P_CO2) carotid body perturbations of 400, 200, 100, 60 and 40 Torr P_O2 were applied in random order, while the brainstem P_CO2 was maintained at 25, 35 or 50 Torr. For the last minute of each 5-min perturbation, the phrenic burst frequency, neural tidal volume and neural minute ventilation were calculated. The effects of brainstem P_CO2 on specific carotid body responses over the hyperoxic (400 to 100 Torr P_O2) and hypoxic (100 to 40 Torr P_O2) ranges were compared using a 2-factor ANOVA (factor 1: brainstem P_CO2, factor 2: carotid body hyperoxia or hypoxia) and the Student-Neuman-Keuls post-hoc test. We found that the phrenic responses resulting from carotid body oxygen perturbations were largest in magnitude when the brainstem was held at 25 Torr P_CO2. This inverse relationship between oxygen responses and brainstem P_CO2 was dose-dependent, suggesting that a hypo-additive (i.e. negative) interaction exists between the carotid bodies and the central chemoreceptors within the respiratory network. This is in contrast to the additive model (i.e. no interaction) of the chemical control of breathing that persists in the human literature.