Legal practice heavily depends on the construction and evaluation of narrative accounts. The ability to legitimately narrate a series of events is a source of power that is unequally distributed in the courtroom. Grounded by a detailed empirical analysis of the court transcripts from R. v. Ghomeshi (2016), this dissertation investigates the relations of power and taken-for-granted assumptions that condition struggles over narrative construction in the sexual assault trial. The project contributes to feminist critiques of sexual assault trials by mobilizing the work of Pierre Bourdieu, which has been largely overlooked in feminist socio-legal scholarship, and by showing how the concepts of narrative capital and what I term configurational power can help us examine narrative power structures. Briefly, narrative capital refers to the speaking positions and properties that bestow authority on one’s narrative practices. The term configurational power refers to the ability to legitimately organize a set of events and experiences into a narrative whole. Through consideration of the conditions and premises that structure who can narrate, in what manner and with what legitimacy, we can better understand the factors contributing to the discrediting of certain testimonies in the courtroom. Analysing the court transcripts revealed several techniques through which the lawyers exercised configurational power and narrative domination over the complainants during the trial: disconnecting and interrupting the complainants’ accounts; highlighting the complainants’ position as unknowing characters; configuring inconsistencies in their accounts; and controlling the narrative ending. Unequal distributions of configurational power constituted a relation of domination that existed as self-evidently legitimate, a form of domination that Bourdieu refers to as symbolic violence. The standards of legal impartiality, autonomy, and objectivity, as well as cultural stock stories about sexual assault and law’s taken-for-granted view of reliable memory, were enacted in the courtroom narrative practices and contributed to the reproduction of this symbolic violence. The unequal relations of narrative capital and configurational power in the courtroom limited the complainants’ ability to narrate their victimization and allowed the defence lawyers to create narrative twists during cross-examination that framed the complainants as manipulative women and upended their claims of victimhood. Through this dissertation, I critically analyze the relations of domination that both condition and were reproduced in the courtroom practices of narrative telling and interpretation.