Parents frequently use contingencies in attempts to regulate children''s food intake. To investigate the effects of instrumental eating on food preferences, each of 45 preschool children was assigned to either an instrumental eating or a control condition. In the instrumental conditions (N = 31), children consumed an initially novel beverage to obtain a reward. To test predictions regarding the contributions of an extension of the response deprivation theory of instrumental performance and extrinsic motivation theory in accounting for negative shifts in preference noted in a previous experiment, 4 instrumental eating conditions were generated by crossing 2 levels of relative amount consumed (baseline, baseline plus) with 2 levels of type of reward (tangible, verbal praise). To control for the effects of exposure on preference, 2 groups of children (N = 7/group) received the same number of snack sessions but with no contingency in effect. Preference data obtained before and after the series of snack sessions demonstrated a significant negative shift in preference for the 4 instrumental groups, while the control groups showed a slight but not significant increase in preference. The implications of the data for child feeding practices are discussed.