Two methods of changing the strength of implicit associations between males and science relative to females and science were evaluated in this study. Participants in each of two experimental conditions completed a manipulation in which they sorted photos of students into sciences and arts according to their stated major. One condition exposed participants to equal numbers of stereotypic and counter-stereotypic images, and the other exposed participants to predominantly counter-stereotypic images. Participants' bias was measured with an IAT pre­ manipulation, post-manipulation, and twenty-four hours later. No significant changes in IAT scores were observed from pre- to post-manipulation. Reasons for the results are discussed, including the nature of the stereotype and methodological considerations. Finally, future research directions and implications for career decisions are considered.