Choosing what to eat is a complex activity. It can require combining visual information about which foods are available at a given time with knowledge of the foods’ palatability, texture, fat content, and other nutritional information. It has been suggested that humans have an implicit knowledge of a food’s fat content; Toepel et al. (2009) showed modulations in visual-evoked potentials after participants viewed images in three categories: high-fat foods (HF), low-fat foods (LF) and non-food items (NF). We tested for behavioural effects of this implicit knowledge. HF, LF, or NF images were used to exogenously direct attention to either the left or right side of a monitor. Then a target (a small dot) was presented either above or below the midline of the monitor, and participants made speeded orientation discrimination responses (up vs. down) to these targets. We found that RTs were faster when otherwise non-predictive HF rather than either LF or NF images were presented, even though the images were orthogonal to the task. These results suggest that we have an implicit knowledge of the fat/caloric/energy value of foods. Furthermore, it appears that the energy benefit of food is present prior to its consumption, after only seeing its image.