This study examined the impact of subtle versus blatant presentation of stereotype threat cues. Subtle and blatant cues were hypothesized to affect performance for different reasons and to emerge on different measures. Subtle cues were hypothesized to cause individuals under stereotype threat to focus resources on reducing uncertainty about bias. To the degree that this causes attention to be directed away from the task, performance on skills requiring concentration and accuracy should be harmed. It was hypothesized that blatant cues, in contrast, do not require uncertainty reduction since bias is clearly present in the environment. Performance can still be negatively affected, however, if individuals utilize strategies that disrupt proper skill execution in an attempt to avoid mistakes that might confirm a stereotype. If so, the impact of subtle versus blatant threat might be revealed on different measures. To test these ideas, female undergraduates who were novel golfers were greeted by a male or female experimenter (subtle stereotype activation) who instructed them that they would complete a task based on golf. Following this, one third of the students were told that the test identifies "psychological factors correlated with general sports performance" (control condition). The other students were told that the task measured "personal factors correlated with natural athletic ability" and half were additionally told that "there are gender differences in sports performance" (blatant gender) and half were told "there are racial differences in sports performance" (blatant race). All students then completed a golf putting task in which the aim was to hit a golf ball into the smallest of three holes on an inclined surface in as few strokes as possible. In terms of the number of strokes needed to complete the putting task, students performed worse only when gender stereotypes were blatantly invoked. Analysis of accuracy on the task, as defined by the hole in which they succeeded in hitting the ball, showed only an effect of subtle stereotype activation; the golfers were less accurate with a male compared with a female experimenter. These data suggest that blatant stereotype threat, even if the source is an ingroup member, can induce a focus on avoiding mistakes, and this focus tends to interrupt the fluid processes that assist performance. Subtle cues, in contrast, are believed to primarily create distractions that impose a cognitive load, minimizing the availability of working memory resources that are needed to concentrate on tasks requiring accuracy. These results suggest that both blatant and subtle forms of stereotype threat can affect performance, but they appear to do so for different reasons and through different mechanisms.