Participants in a selection interview, especially applicants, are highly motivated to display information that is to their advantage and will deploy impression management tactics. Yet, recruiters have trouble detecting such tactics. One possible explanation is that some tactics may be more subtle and less obvious to perceive, and thus are unintentionally neglected. This dissertation aims to offer a glimpse of some relevant but little-known interactional aspects that participants deploy during a selection interview and to explore their impact on interview outcomes. I explored three phenomena: disfluencies, storytelling and laughter. First, I explored the applicants’ responses to past-behavior questions. These questions, from an interactional point of view, require applicants to respond in a specific, narrative format. In article 1, one specific type of disfluencies, response delays, predicted applicants’ response type. As time went by, applicants were less likely to produce the appropriate narrative response. Response delays also negatively predicted recruiters’ hiring evaluations. Responding in time and with the appropriate response type has positive outcomes for applicants in terms of selfpresentation goals. Results also pointed out the difficulties applicants experienced to produce narratives on demand. In article 2, I manipulated two ways of enhancing the applicants’ storytelling: probing and information. Recruiters’ interactive behavior (probing) increased the applicants’ production of stories and the variety of narrative elements they contained. Recruiters have a key role to play to enhance the applicants’ storytelling. The applicants’ level of information did not increase the production of stories. However, information decreased the production of pseudostories. Additionally, results confirmed that finding a relevant episode to narrate was a major issue for applicants. Finally, I explored a rather original but promising interactional aspect, laughter (Article 3). In the selection interview, laughter is more likely to be produced by applicants (compared to recruiters), by women and on transitions (between different phases of the interview). Furthermore, shared laughter episodes initiated by recruiters were positively related to their hiring evaluations, whereas applicants’ unilateral laughter was negatively related to recruiters’ hiring evaluations. Laughter is not produced randomly and constitutes a subtle way to manage impressions in the selection interview. Key results and implications for selection interviews, limitations and future research perspectives are discussed.