The widely used paradigm developed by Katz and Braly (1933) for the assessment of “stereotypes” has been criticized because of the ambiguity of the subject's task. In the present study, 2 alternate sets of instructions were used, attempting to make the subject's task more clearcut. Traits attributed to each of 5 ethnic groups were virtually identical to those given under the instructional format usually utilized in such studies. When subjects were asked to list the trait-attributions most commonly made by “others” in our society to each of the ethnic groups, responses concerning Negroes were considerably different from subjects' own attributions to Negroes. This own-others difference was not evident for the other 4 ethnic groups rated. Racial attitude was significantly related to attribution for 5 of the 17 traits attributed to blacks by over 10% of the subjects, when subjects were giving their own views. The degree to which a subject's attributions to Negroes agreed with those of the sample as a whole was not related to his racial attitude. However, the degree to which a subject's attributions to Negroes agreed with those seen as most commonly attributed by “others” in our society was positively related to racial prejudice. The implications of these findings for common conceptualizations of stereotype are discussed.