The memory-improving action of post-training, noncontingent injections of glucose was investigated in a series of experiments which examined the effects of several substances that interact with glucose metabolism on the retention of a conditioned emotional response and on blood glucose levels in male hooded rats. Although post-training glucose injections of 1, 2, and 3 g/kg all produced similar increases in blood glucose, only 2 g/kg improved retention, suggesting that attainment of a particular blood glucose level is not critical for memory improvement. Post-training injections of a range of insulin doses (0.25-4 IU/kg) failed to affect retention. Post-training injection of fructose (the same doses as were used for glucose) had no effect on blood glucose levels and, as with glucose, only the 2 g/kg dose improved retention. This finding suggests that blood glucose levels are not critical for the memory-improving effect, that glucose and fructose may act on the same substrate and, because fructose does not act directly on the brain, it raises the possibility that both substances act peripherally. Post-training injections of 2-deoxyglucose and 3-O-methylglucose both improved retention. The fact that these mostly nonmetabolized glucose analogs were effective suggests that the memory-improving action of glucose may depend on the activation of a membrane glucose transport mechanism. The implications of the possible action of glucose on peripheral transport mechanisms for understanding the effect of reinforcers on memory are discussed.