Two experiments investigated conditions under which stereotype threat might produce reactance, behavior in opposition to stereotype threat. In Experiment 1, male-female negotiation dyads were given either a strong or weak gender stereotype manipulation implying male superiority in negotiating. One of the negotiation partners was provided with social power by suggesting that they could leave the negotiation and still receive a good settlement, and the individual with high power was female in half the dyads and male in the other half of dyads. Results showed that high-powered negotiators were more effective than low-powered negotiators, but this was true only when male gender stereotypes had been strongly instantiated. In addition, negotiators behaved in a more contentious fashion under the strong stereotype manipulation. These effects occurred regardless of gender, suggesting that women do not succumb to, but in fact reacted against, stereotype threat. In Experiment 2, stereotypes were either strongly or weakly invoked to suggest female superiority in negotiation. High-powered negotiators performed more effectively than low-powered negotiators, but joint payoffs were maximized in the strong female stereotype condition, suggesting that female stereotypes of generosity, attentiveness, and communication prevailed for both male and female negotiators in this condition. These results show that when stereotypes are activated explicitly, individuals low in power who are subjected to the limiting stereotype can assimilate their behavior to attributes that are tied to success in a specific domain.