The ketogenic diet is an effective treatment for seizures, but the mechanism of action is unknown. It is uncertain whether the anti-epileptic effect presupposes ketosis, or whether the restriction of calories and/or carbohydrate might be sufficient. We found that a relatively brief (24 h) period of low glucose and low calorie intake significantly attenuated the severity of seizures in young Sprague-Dawley rats (50-70 gms) in whom convulsions were induced by administration of pentylenetetrazole (PTZ). The blood glucose concentration was lower in animals that received less dietary glucose, but the brain glucose level did not differ from control blood [3-OH-butyrate] tended to be higher in blood, but not in brain, of animals on a low-glucose intake. The concentration in brain of glutamine increased and that of alanine declined significantly with low-glucose intake. The blood alanine level fell more than that of brain alanine, resulting in a marked increase ( approximately 50%) in the brain:blood ratio for alanine. In contrast, the brain:blood ratio for leucine declined by about 35% in the low-glucose group. When animals received [1-(13)C]glucose, a metabolic precursor of alanine, the appearance of (13)C in alanine and glutamine increased significantly relative to control. The brain:blood ratio for [(13)C]alanine exceeded 1, indicating that the alanine must have been formed in brain and not transported from blood. The elevated brain(alanine):blood(alanine) could mean that a component of the anti-epileptic effect of low carbohydrate intake is release of alanine from brain-to-blood, in the process abetting the disposal of glutamate, excess levels of which in the synaptic cleft would contribute to the development of seizures.