Women remain underrepresented in the fields of science, technology, engineering, and mathematics (STEM). While other research focuses on personal factors (e.g., group identification, stereotype endorsement) responsible for this underrepresentation, the present thesis focuses on understanding how the cues women are exposed to in mathematics and science situations also contribute to the lack of women in STEM. Specifically, women may withdraw motivation to pursue STEM careers because the "situational cues" they are exposed to in these career settings lead them to feel uncertain as to how well they will belong and perform. Accordingly, women's sense of belonging and expected performance if they were to take a hypothetical mathematics test were measured in response to various situational cues manipulated within a fictitious report about the current state of women in STEM. First, group composition cues (i.e., cues that suggest women may or may not be represented equally in STEM) were examined. Past research shows that when presented with unbalanced composition cues, women feel less belonging and perform worse compared to when presented with balanced composition cues. However, this research has always examined group composition cues in relation to relevant comparison groups (e.g., students attending the participant's university). Therefore, it is still unclear whether women are affected by composition cues when presented in relation to less relevant comparison groups (e.g., individuals in the national workforce). This is particularly problematic because women are underrepresented in STEM at these national levels and are thus likely to experience these "less relevant" composition cues in their daily lives. Therefore, this study investigated the impact of group composition cues while also varying the relevance of these cues. It was hypothesized that group composition cues would have less of an impact on college women's perceptions when presented in relation to less relevant comparison groups because these groups are not as diagnostic or informative about the experiences of college women. To test these predictions, 162 undergraduate women at San Diego State University were presented with these cues in an online session lasting approximately 15 min. Results showed that participant's sense of belonging in STEM was dependent on the relevance of the situational cues. When exposed to relevant comparison groups, participants reported a greater sense of belonging than when presented with less relevant comparison groups. While this effect did not depend on group composition, it is important that it was only in the relevant condition (SDSU) that participants felt more belonging following the balanced as opposed to the unbalanced composition cues. When exposed to less relevant composition cues, this pattern actually reversed. Furthermore, there was an interaction of composition and relevance on participants' expected mathematics performance. Participants exposed to relevant composition cues showed higher expected mathematics performance following a balanced as opposed to an unbalanced group composition. Participants exposed to less relevant comparison groups showed no difference as a function of group composition. These findings indicate that relevance impacts how women are affected by group composition cues and demonstrates the need for further studies to explore multiple cues and cue properties.