In this realistic experiment, an interview with the leader of the Liberals in the Dutch Parliament was recorded in the presence of a live audience, which reacted in a positive, negative, or neutral way. It was shown to subjects of two opposing political parties, whose attitudes were to be changed by the experimental interview. The main hypothesis, which predicted more attitude change in the positive than in the negative audience condition, could not in general be supported. The alternative audience attraction hypothesis was mainly sustained: With an audience, perceived as attractive, attitude change was greatest when the audience reacted positively and least when it reacted negatively, while for an unattractive audience the opposite effect was demonstrated.