Two studies examined the effect of gender stereotypes on students' recollection of their school marks in stereotypically feminine (arts) and masculine (mathematics) domains. As predicted, the results of Study 1 indicated that the more students believed in gender stereotypes prior to recall, the more they biased their reported marks, compared to their actual marks, in a stereotype-consistent way (female students underestimated their marks in mathematics and male students underestimated their marks in arts). Study 2, in which the salience of gender stereotypes was manipulated prior to recall, yielded similar findings. The recall of school marks was more stereotype-consistent in a condition of high salience than in a condition of low salience of gender stereotypes. The theoretical implications of these results are discussed.