Individuals exhibiting restrained eating behaviors demonstrate increased inhibitory control when exposed to food-related stimuli, indicating the presence of an automatic food-inhibition association. Existing literature proposes that this association contributes to the devaluation of food within this population. Efforts to disrupt this association by promoting the complete elimination of the inhibition of food responses have resulted in increased food consumption but have also led to heightened food-related anxiety in individuals with restrained eating behaviors. In the current investigation, we investigated whether a novel flexible food response/inhibition computerized task could yield favorable changes in attitudes toward food in individuals with restrained eating. We randomly assigned 78 females who engage in restrained eating to one of three training groups. In the flexible response/inhibition group, participants were instructed to equally inhibit or respond to food stimuli. In the response group, participants consistently responded to food stimuli, while in the inhibition group, participants consistently inhibited their response to food cues. Implicit attitudes toward food were assessed both before and after the manipulation. To examine the stability of the effect of the training, participants also engaged in a seemingly unrelated bogus taste test. Our results revealed that only the flexible response/inhibition group demonstrated a significant improvement in positive attitudes toward high-calorie foods after eating, while there were no observable changes in negative attitudes among the other two groups. These findings suggest that promoting a balance between the responding and inhibiting responses to food stimuli can increase positive attitudes toward food amongst individuals with restrained eating.