The Implicit Association Test (IAT; G. Greenwald et al; see record 1998-02892-004) provides a flexible measure of the automatic associations underlying implicit prejudice. Results of three experiments showed strong evidence of implicit prejudices based on religious ethnicity (Jewish vs Christian), age (young vs old), and nationality (American vs Soviet). Subjects responded more rapidly to tasks that obliged association of ingroup tokens to pleasant attributes and outgroup tokens to unpleasant attributes than to ones that obliged the complementary associations. In addition, the findings of three experiments were consistent with the hypothesis that IAT effects are independent of self-reported stimulus familiarity differences. These results support the construct validity and the generality of the IAT method in implicit prejudice research.