This paper investigated misattribution and self-affirmation as means for reducing stereotype threat. Study 1 replicated typical stereotype threat effects with students (N = 93; 30 men, 63 women) who completed a math test after either being exposed to information suggesting either that the test produced gender differences (stereotype threat for women) or was gender fair (control). Findings showed that women performed significantly worse on the math task than men in the stereotype threat condition. In addition, women in the stereotype threat condition were less confident in their answers than were men in the same condition, and differences in confidence appeared to mediate the relation between stereotype threat and performance. Study 2 employed similar procedures as Study 1 with female undergraduates (N = 42) who also were randomly assigned to one of four experimental conditions created by crossing misattribution (present versus absent) and self-affirmation (self-affirmation present versus absent) variables. The misattribution factor was manipulated by the presence or absence of a malfunctioning computer monitor during performance that several previous participants had said made them feel anxious. Self-affirmation was manipulated by having half of the participants write an essay comparing their academic achievements with a student with a middling academic profile. Results showed that women women in the self-affirmation present condition performed better on the test than did participants who did not have a chance to self-affirm. There was also some evidence that the misattribution manipulation improved test performance. Consequently, women in the self-affirmation/misattribution present condition produced the highest test scores. These results provide evidence that self-affirmation, especially when combined with opportunities to misattribute feelings of anxiety to external circumstances can reduce the negative consequences of stereotype threat.