Research interest in body image has accelerated in recent decades. As such, there exists a substantial body of literature examining affective and behavioural correlates of body dissatisfaction. Although this literature has consistently demonstrated positive relationships between body dissatisfaction and depression, social physique anxiety, disordered eating and physical inactivity, the underlying mechanisms of the relationships between one' s body image and affect remain poorly understood. Furthermore, although body image researchers have c01mnonly examined discrepancies between one' s current (or actual) self and one' s ideal self, they have not commonly done so within an established theoretical framework. Higgins' (1987) self-discre·pancy theory proposes that individuals possess three selves: an actual self, an ideal self, and an ought self; and that discrepancies between these selves result in distinct emotional consequences. Furthermore, it has been proposed that individuals possess a feared self, which may act as a moderator of the relationships between ideal and ought selves, and affect. Although this proposal has begun to receive research attention within the global self-discrepancy literature, it has yet to be examined within the context of body image. As such, utilising a self-discrepancy theory framework, this thesis presents a detailed examination of ideal, ought and feared body image self-discrepancies and their interactions as predictors of social physique anxiety, body dissatisfaction and disordered eating attitudes. The first study extends the work of Carver et al. ( 1999) and Woodman and Hemmings (2008) by examining the moderating role of the feared self in the relationship between women' s ought body fat discrepancies and social physique anxiety. Specifically, we examined the hypothesis that the positive relationship between ought body fat discrepancies and social physique anxiety will be stronger when women are farther from their feared self. Results were in line with that hypothesised and provide initial support for such an interaction framework in examining body self-discrepancies; highlighting the importance of considering the role of the feared self in the relationship between body self-discrepancies and social physique anxiety. Study 2 was an examination of the interaction between ought and feared body fat discrepancies in predicting social physique anxiety and disordered eating attitudes. Study 2 extended the findings of Study 1 by examining a model of mediated moderation such that ought and feared discrepancies interact to significantly predict disordered eating attitudes, and that this relationship is mediated by social physique anxiety. Results demonstrated that the ought x feared interaction on disordered eating attitudes was fully mediated by social physique anxiety. However, the fonn of the ought x feared interaction was in contrast to that hypothesised (cf. Study 1). Specifically, the positive relationship between ought body fat discrepancies and disordered eating attitudes was significant only when women were close to their feared self. Study 3 addressed the growing body of correlational and experimental literature examining the effects of acute exposure to media ideals on body image and associated affect (e.g., Halliwell & Dittmar, 2004, 2005). We extended this by examining the moderating role of exposure to media ideals on the relationship between body discrepancy magnitude and affect. Specifically, we examined whether the positive relationships between ideal discrepancies and body dissatisfaction; and ought body discrepancies and social physique anxiety, are attenuated by exposure to media ideals. Results provided support for these hypotheses such that the positive relationship between ideal body discrepancies & body satisfaction, and ought body discrepancies & social physique anxiety, were weaker following exposure to media ideals compared to a control condition. Utilising a hierarchical linear modelling approach, Study 4 examjned within-person variability in the specific shape of interactions between self-guides in predicting social physique anxiety. Specifically, we examined whether the moderating role of the feared self differs as a function of mean levels of social physique anxiety. Results demonstrated a three-way cross-level interaction such that when mean social physique anxiety was high, the positive relationship between ideal discrepancies and social physique anxiety was significant only when participants were relatively far from their feared self. In contrast, when mean social physique anxiety was low, the positive relationship between ideal discrepancies and social physique anxiety was significant only when participants were relatively close to their feared self.