Previous research has consistently found enhancement of memory after the ingestion of a glucose containing drink. The aims of the present study were to specify more precisely the nature of this facilitation by examining the cognitive demand hypothesis. This hypothesis predicts greater glucose induced facilitation on tasks that require significant mental effort. In two experiments, both employing an unrelated sample design, participants consumed either 25g of glucose or a control solution. In experiment 1, participants first studied low and high imagery word-pairs and memory was assessed 1-, 7- and 14-days later by cued recall. Overall, glucose enhanced both encoding and consolidation processes only for the more difficult low imagery pairs. In experiment 2, the degree of mental effort in a verbal memory task was manipulated in two ways: (1) by varying the phonological similarity of the words; and (2) by varying the length of word lists. Glucose was found to enhance memory only for longer word lists. These data are consistent with the idea that glucose is especially effective in demanding memory tasks, but place some limits on the forms of difficulty that are susceptible to enhancement.