The present study investigated the extent of expectancy in the ability of glucose to affect cognitive performance. Using a within-subjects design, subjects (n 26) completed four experimental sessions (in counterbalanced order and after an initial practice session) during which they were given a 500 ml drink 30 min prior to completing a cognitive assessment battery. In addition, all subjects completed a baseline practice session during which they were given no drink. During two of the sessions, subjects were given a drink containing 50 g glucose and on the other two they were given a drink containing aspartame. A balanced placebo design was used, such that for half the sessions subjects were accurately informed as to the content of the drink (glucose or aspartame), whereas in the other two sessions they were misinformed as to the content of the drink. The task battery comprised a 6 min visual analogue of the Bakan vigilance task, an immediate verbal free-recall task, an immediate verbal recognition memory task and a measure of motor speed (two-finger tapping). Blood glucose and self-reported mood were also recorded at several time points during each session. Glucose administration was found to improve recognition memory times, in direct contrast to previous findings in the literature. Glucose administration also improved performance on the Bakan task (relative to the control drink), but only in sessions where subjects were informed that they would receive glucose and not when they were told that they would receive aspartame. There were no effects either of the nature of the drink or expectancy on the other measures. These results are interpreted in terms of there being some contribution of expectancy concerning the positive effects of glucose on cognition in studies which have not used an equi-sweet dose of aspartame as a control drink.