Although glucose typically enhances memory or reverses memory deficits, glucose can also produce memory deficits when co-infused with the gamma-aminobutyric acid (GABA) agonist muscimol into the medial septum (Parent & Gold, 1997; Parent, Laurey, Wilkniss, & Gold, 1997). To date the mechanisms underlying the memory-impairing interaction between GABA and glucose remain unknown. Here we investigate whether this effect is the result of hyperosmolar conditions or may involve glucose metabolism. Male Sprague-Dawley rats were given one-trial inhibitory avoidance training after receiving septal infusions of vehicle (phosphate-buffered saline, 0.5 microl), the GABA(A) agonist muscimol (3 nmol), glucose (16.5, 33, or 66 nmol), fructose (33 nmol), pyruvate (33 nmol), or a solution containing muscimol combined with glucose, fructose, or pyruvate. Retention performance was tested 48 h later. Infusions of glucose, pyruvate, fructose, or muscimol alone did not affect retention performance. However, co-infusions of all doses of glucose (16.5, 33, or 66 nmol) or the glycolytic end product pyruvate with muscimol impaired retention performance. Co-infusions of fructose with muscimol did not affect retention performance. These results suggest that the memory-impairing interaction between glucose and muscimol does not result from hyperosmolar conditions, because equiosmolar concentrations of fructose do not mimic the effects of glucose and the memory deficits do not vary as a function of glucose concentration. The finding that pyruvate mimicked the effects of glucose and impaired memory when combined with muscimol suggests that glucose metabolism may be involved in the memory-impairing interaction between glucose and GABA(A) receptors in the medial septum.