Subjects possessing two distinct types of self-discrepancies, actual:ideal (AI) and actual:ought (AO), read a persuasive message about the importance of eating breakfast, framed in terms of either positive or negative outcomes. On the basis of an analysis of each discrepancy type as a chronic, individual motivational force and each frame as a momentary, situational motivational force, the positive outcome frame was predicted to be more effective than the negative outcome frame in motivating AO subjects to change their eating patterns; the opposite was predicted for AI subjects. The results supported this prediction. AO subjects' thoughts, feelings, and intentions showed a stronger persuasive effect of positive outcome framing than of negative outcome framing; the opposite was true for AI subjects. The predicted interaction was also found on an immediate behavioral commitment measure. On a delayed action measure, only the effect predicted for AO subjects was found.