This study examines the ability of the Theory of Planned Behavior to account for mothers' intentions to limit the frequency of their infants' sugar consumption. One group of mothers (experimental) was exposed to a dental health education program designed to promote this behavior, the other (control) group was not. Each group of mothers was interviewed twice; the intervention was delivered to mothers in the experimental group immediately after their first interview. The interviews incorporated measures of the key constructs in the Theory of Planned Behavior (attitudes to behavior, subjective norms, perceived behavioral control, and behavioral intentions). Results indicated that the addition of perceived behavioral control resulted in small but significant increments in the amount of explained variance in intentions, thereby supporting the theory. Although the change in intentions to perform the advocated behavior was not significantly greater in the experimental group than in the control group, there was a significant change in attitudes to the behavior in the experimental group but no such change in the control group. Moreover, amount of change in behavioral intentions was significantly correlated with amount of change in attitudes, but not with amount of change in subjective norms or perceived behavioral control. Analysis of behavioral beliefs indicated that the attitude change reflected a change in a specific behavioral belief that was targeted in the intervention. Further analyses focused on the role played by direct experience of the behavior and show that such experience tended to enhance the role of perceived behavioral control in intention formation, apparently because mothers who have older children have discovered that they have relatively little control over frequency of child's sugar consumption. The implication of these results for the Theory of Planned Behavior are discussed.