Social values are ideals that people consider to be important guiding principles in their lives (e.g., freedom, equality). This presentation will examine the cognitive and affective components of social values and how these components help to predict value-relevant behavior. One series of experiments used different paradigms to examine the cognitive component. These experiments revealed that most values function as cultural truisms (i.e., beliefs that are widely shared and rarely questioned), because beliefs about values are relatively inaccessible during value judgments, and access of such beliefs actually causes value change. A second series of experiments showed that this lack of cognitive support for values is important, because tasks encouraging the development of new value-relevant beliefs elicit value-congruent behavior in situations that make it difficult to perform such behavior. Moreover, this effect occurs primarily when the elicited beliefs are concrete and refer to prototypical situations encompassing the values. Finally, a third series of experiments explored the affective component of values, finding that values elicit qualitatively different emotions depending on their roles as self-guides. Together, this evidence helps us to better understand how social values are represented in memory and guide relevant behavior.