We investigated effects of coffee cue presentation on desire for coffee and cognitive performance. The 2 (cue and no-cue)×2 (instruction: reward and no-reward) between-subjects design was used. The smell and sight of coffee were presented in the cue condition, but not in the no-cue condition. The participants in the reward condition were instructed that they would obtain coffee after the behavioral task and the amount of coffee depended on their performance of the task. The participants in the no-reward condition were instructed to perform as many tasks as possible. The dependent variable was performance of the behavioral task and subjective desire for coffee. In the task, the participants were asked to find vowels among letters printed on task sheets. As a result, the participants in the cue condition found more vowels than those in the no-cue condition, in both instruction conditions. There was no difference in subjective rating between any conditions. These results suggest that the coffee cue may enhance cognitive performance rather than desire for coffee.