In this reply, I criticize Bartsch and Judd's (1993) article on several grounds. First, they under-utilize the efforts undertaken in prior work to rule out the possibility of an inverse relation between group size and perceived group homogeneity as an alternative explanation of the observed ingroup homogeneity effect. Secondly, Bartsch and Judd's design doubles and thus aggravates the confounding problem. By trying to avoid the target group size confound, they end up with two other confounds involving level of abstractness and frame of reference. Finally, I criticize Bartsch and Judd's methodological advice to avoid within-subjects comparisons of ingroup and outgroup homogeneity in minority–majority contexts. Quite on the contrary, I highlight the socialpsychological significance of these comparisons.