These studies were undertaken to investigate the relationship between medium-chain fatty acid availability, medium-chain fatty acid oxidation, and central nervous system toxicity during infusion of medium-chain triglycerides in dogs. Six dogs received a sequential, stepwise infusion of trioctanoin at three different rates for 80 min each, providing calories below and equal to resting energy expenditure in the species. Ketone body production rates (using a 14C beta-hydroxybutyrate tracer) and plasma concentrations of lactate and octanoate were monitored. Three animals were infused with saline to serve as controls. Blood-brain barrier integrity was assessed with Evans blue dye, and brain samples were taken at the end of the study to quantify brain water. Three animals were studied under anesthesia to obtain good quality EEG and intracranial pressure measurements. Results were (1) plasma octanoate increased to 0.37 +/- 0.13, 0.78 +/- 0.2, and 1.44 +/- 0.41 mmol/liter during the three infusion intervals; (2) emesis, somnolence, and coma were observed at the two highest trioctanoin rates; (3) ketone body concentrations and production increased from 102 +/- 15 to 859 +/- 54 mumol/liter and 3.6 +/- 0.43 to 18.5 +/- 1.7 mumol/kg/min, respectively, at the highest trioctanoin infusion rate; and (4) plasma lactate also increased from 1.3 +/- 0.1 to 4.3 +/- 0.9 mmol/liter at the highest infusion rate. EEG changes were also observed, consisting of high amplitude slowing and reduction in amplitude of faster components. There was no extravasation of Evans blue dye, nor change in brain water or intracranial pressure. The conclusion--medium-chain triglycerides have significant dose-related central nervous system toxicity in dogs. Therefore, caution should be exercised in clinical studies with MCTs, including careful measurement of medium-chain fatty acid concentrations.