The research examined an additional proposed moderator of the attitude-behavior relationship: the activity level of emotions associated with an attitude object. In Experiment 1, participants who self-generated active rather than passive emotions as being associated with gay men displayed greater attitude-behavior consistency in hiring recommendations for a gay job applicant, as did participants who rated active rather than passive experimenter-provided emotions as being associated with gay men. In Experiment 2, participants who were instructed to associate active rather than passive emotions with gay men subsequently displayed greater attitude-behavior consistency in similar hiring recommendations. It is suggested that future research on the affective component of attitudes might benefit from going beyond consideration of whether the associated emotions entail displeasure or pleasure.