Two experiments are reported that used the affective priming paradigm (Fazio, R. H., Sanbonmatsu, D. M., Powell, M. C., & Kardess, F. R. (1986). On the automatic activation of attitudes. Journal of Personality and Social Psychology, 50, 229-238) to uncover associations with food at a relatively automatic level. Experiment 1 tested the hypothesis that anorexia nervosa (AN; n=22) patients would show less sensitivity to the palatability of foods than unrestrained lean controls (n=27). Results indeed suggested that AN patients did not display a liking of palatable foods over unpalatable foods, whereas unrestrained controls did. Experiment 2 tested the hypothesis that obese people (n=27) would show more sensitivity to the palatability of (high-fat) palatable foods than unrestrained lean controls (n=27) would. However, results suggested that the priming effect was based on health concerns, in that participants showed a preference for low-fat palatable foods over high-fat palatable foods. Average speed of responding and context are discussed as variables influencing the affective priming effect. Taken together, results suggest that food evaluations at a relatively automatic level are controlled by an interaction between participant characteristics, stimuli characteristics, and the specific context.