A review of 80 years of writing and research on stereotypes indicates that there is little agreement on a definition of stereotype; also that the major emphasis has been on definitions that emphasize the consensus of a group on the elements of a stereotype. This paper defines stereotypes in terms of characteristics ascribed by individuals, rather than groups. It suggests a method of measuring "extent of stereotype" which depends on two factors: the number of characteristics used in the stereotype (e.g., stupid, cheerful), and the intensity with which a characteristic is attributed (e.g., all X are stupid, some Y cheerful). Three versions of an instrument developed to measure stereotype, using a modified version of Veldman and Parker's list of personality traits, were administered to 474 high school students, selected only on the basis of providing maximum variability of responses. Results of the analysis, not yet complete, support the validity and stability of the instruments and indicate a moderate degree of correlation (.34 to .70) among the subjects' holding of stereotypes with respect to women, police, and people in general. (RM)