This research investigated in two experiments the relationship between acquiring academic knowledge and stereotype threat. Experiment 1 was conducted with students (N = 76; 32 Black, 44 White) who, in an initial session, were asked to learn some new information to determine either how well people from different backgrounds learn (stereotype threat for Black students) or psychological factors that contribute to different learning styles (control). They then were provided with a list of 24 novel words with their definitions (e.g., succedaneum) and given 10 minutes to learn the words. From one to two weeks later, participants returned and completed several tests of their retained knowledge of the words. First, they completed a warm up task involving recalling and matching the definitions of 12 of the 24 words. This task was intended to offer a relatively low a pressure and low stereotype threat testing situation. Next, participants were told they would complete two tests that would evaluate your ability to learn verbal information and your performance on problems requiring verbal reasoning ability. These instructions were designed to elicit high pressure and high stereotype threat by highlighting negative relevant stereotypes about African Americans. Participants then recalled and matched the definitions for the remaining 12 words. Overall, memory performance in the low pressure testing phase showed that Black students performed worse if they had learned under stereotype threat compared with control instructions. White students, in contrast, were generally unaffected by the threat manipulation present at learning. In the high pressure testing situation, performance of Black students was harmed to a greater degree than the performance of White students. However, the effects of stereotype threat at learning remained clear in Black but not White participants. These results suggest that stereotype threat encountered in either the learning or performance environments can affect the ability to learn and recall new information. Experiment 2 involved Black students (N = 36) who were all given the identical stereotype threat instructions from Experiment 1 but, prior to studying the words, completed a value affirmation writing manipulation. Specifically, they either were asked to select a value from a set and write why it was important to them (affirmation condition) or to select their least important value and why it might be important to someone else (no affirmation condition). Later memory performance indicated that students who had self-affirmed in the learning environment performed better than students who had not self-affirmed, and this was particularly in the low pressure testing situation. Additional measures suggested that the self-affirmation manipulation improved performance by preventing attempts to suppress stereotypes (a cognitively-taxing process) and by encouraging a promotion focus, a focus on achieving successes rather than avoiding failures. These results show that stereotype threat can undermine the ability to learn new materials and to recall that material in high pressure testing situations, but that these negative effects can be avoided when learners feel self-affirmed.