AbstractWith no theory and the use of different methods and assumptions, Coenders and his colleagues not surprisingly produce a different analysis of prejudice from ours of 1995. But we see no reason to withdraw our claims for both the concept and measurement of subtle prejudice. By obscuring our theoretical structure and offering none of their own, our critics follow their empirical preferences to reach a meaningless solution. In doing so, they disregard the social psychological literature on prejudice—especially that of Rokeach and Tajfel. Moreover, their reanalysis of the 1988 Eurobarometer survey data does not replicate our analysis in important ways. We answer their objections to our use of orthogonal varimax rotation and our demonstrations of the predictive value of the Subtle Prejudice Scale, and we show that when tested appropriately our theoretically derived model performs as well as and often significantly better than their atheoretical model. We also find the six conditions set down by Coenders et al. for differentiating the two forms of prejudice are largely met by our analysis. Finally, repeated replications by other investigators in ten countries with diverse outgroups and new samples support both our original analysis and the usefulness of our Blatant and Subtle Prejudice Scales. Copyright © 2001 John Wiley & Sons, Ltd.