Recent theory and research suggest that certain situational factors can harm women's math test performance. The three studies presented here indicate that female role models can buffer women's math test performance from the debilitating effects of these situational factors. In Study 1, women's math test performance was protected when a competent female experimenter (i.e., a female role model) administered the test. Study 2 showed that it was the perception of the female experimenter's math competence, not her physical presence, that safeguarded the math test performance of women. Study 3 revealed that learning about a competent female experimenter buffered women's self-appraised math ability, which in turn led to successful performance on a challenging math test. Fewpeoplewoulddisagreewiththeideathatrolemodels can have a profoundly positive impact on a person's life. Role models may differ in terms of their age, race, and/or sex, yet there is one feature that they share: They are all perceived to be competent in their respective areas (Lockwood & Kunda, 1997). For instance, female role models in math-related domains might be particularly helpful for math-talented women because they represent stereotype-disconfirming evidence about women's inferior math ability, so that women's math test performance is protected after encountering or learning about a female role model. In addition, if a low score on a standardized math test is one of the reasons why women are seriously underrepresented in math and engineering, then the benefits of having female role models for female students in those academic domains may be considerable. GENDER DIFFERENCES IN MATH Over the years, a vast amount of research has been devoted to investigating gender differences in math. Even though much of this work shows that such differences are relatively scarce (Hyde, Fennema, & Lamon, 1990; Kimball, 1989), there is a general pattern to the differences that do exist. Women, for instance, perform as well as men on less advanced math problems, but when the math problems become more advanced, women do not perform as well as men (Gallagher et al., 2000; Halpern, 1992; Kimball, 1989). This difference often does not emerge until high school (Hyde et al., 1990), precisely when performance on standardized math tests-tests that contain more advanced math problems-may matter the most. But perhaps the most far-reaching gender difference is the fact that female students, unlike male students, routinely encounter negative stereotypes about their math ability