People encounter a wide range of objects in the visual environment. Some of these are important and others are unimportant. Given that limitations on both attention and memory constrict people’s ability to process everything in an environment, they must somehow prioritize important over unimportant objects. One suggestion as to how people do this is that the contents of working memory (WM), likely pertinent to current objectives, bias attention to environmental stimuli with which they share features. This suggestion has received mixed support in the literature. The degree to which stimuli in WM bias attentional deployment may relate their task relevance, that is, their pertinence to one’s present activity. However, considering that emotional stimuli appear to command attention preferentially, this thesis asks how the presence of emotional stimuli, specifically expressive faces, bias attentional deployment. Research in the area of emotion-attention interactions suggests that emotions of different valence have distinct effects on attention. Positive emotion leads to a broader, more efficient, allocation of attention than negative emotion. Thus, emotion-related WM-attention interactions may result in distinct patterns of attention capture depending on the valence of the emotion involved. We tested whether this account describes the interaction between the emotion on a face held in WM and visual attention during the performance of a subsequent task involving emotional schematic faces. Consistent with expectations, emotion in WM influenced attention. Specifically, positive emotion led to a broader attentional focus than negative emotion. Importantly, emotion only influenced attention when it was task-relevant (Chapter 2). Event-Related Potential data indicated that when emotion was not task-relevant, participants processed WM-matching expressions more superficially than non-matching expressions, suggesting that WM-matching contents are dismissed more quickly when not task-relevant. Nonetheless, these stimuli interfered with visual processing; a result that may explain observed discrepancies in WM-attention interactions with non-task-relevant stimuli (Chapter 3). Finally, we extend the finding that positive emotion leads to faster target processing, to the concept of value. Here, we examined how the intrinsic value of an emotional expression related to its ability to capture attention. Research shows that people are willing to give up money for the chance to see genuine smiles. Thus, we hypothesized that a genuine smile’s subjective value would predict attention capture for genuine-smile targets in a flanker task. Results confirmed this prediction, suggesting that an expression’s intrinsic value also drives attention capture and may therefore have implications for how people navigate social interactions (Chapter 4). Together, these results suggest that emotion in WM biases attention in a manner that is sensitive to the demands of a current task. Specifically, whereas task-relevant positive emotion results in more efficient orienting of attention than task-relevant negative emotion, non-task-relevant emotion receives demoted priority in visual processing. These results extend to an item’s value, such that higher-value stimuli receive priority processing. This research extends our understanding of WM-attention and value-attention interactions to include long-term semantic associations as a factor. Collectively, the results of this research suggest that the allocation of attention to social stimuli is determined based on social implications, with positive implications having particular influence.