Serial, paired blood samples were collected via cannulae chronically placed in the common carotid artery (A) to and the internal jugular vein (V) from the brain of the fasted adult American bullfrog (Rana catesbeiana). Plasma glucose, beta-hydroxybutyrate, acetoacetate, lactate, and alanine levels were measured by standard enzymatic procedures. Cannula failure ended sampling after 1-2 days in most animals. The common carotid artery plasma metabolite levels were greatest at the time of surgery and subsequently declined to relatively stable levels. The summarized data indicated glucose uptake and alanine release by the brain, but no significant beta-hydroxybutyrate or lactate A-V percentage changes. Initially, acetoacetate levels also were measured, but were discontinued in favor of continued beta-hydroxybutyrate determinations when no significant A-V percentage changes occurred. Separate analysis of the metabolite levels during the surgery and recovery period (less than or equal to 24 hr) and the "normal" under the experimental conditions period (greater than 24 hr) revealed that summarizing the data masked important A-V percentage changes during the two different physiological conditions. Glucose was the only metabolite extracted by the brain during the less than or equal to 24 hr period of elevated and subsequently declining metabolite levels. In contrast, glucose uptake did not occur during the greater than 24 hr period of stable levels, but there was lactate release. If the bullfrog brain stores substantial glycogen as do the other ectothermic vertebrates studied, glucose uptake when plasma levels are elevated, for example after feeding, may serve both to fuel the brain and to replenish endogenous glycogen reserves that may be mobilized to provide glucose for the brain after plasma glucose levels return to normal. Assuming that mammalian and bullfrog metabolic pathways are the same, the release of lactate and alanine by the brain, possibly to remove excess pyruvate and to regenerate NAD+, is consistent with this hypothesis. It remains to be determined for how long endogenous energy sources alone can support the bullfrog brain, and if plasma glucose, ketones, and/or other energy sources are extracted as endogenous brain fuels become exhausted.