Relatively recent self-talk research has implicated the role of grammatical features, such as first and second-person pronouns, as holding relevance for performance. Nevertheless, there is a general lack of data regarding how individual differences (e.g., personality traits) might impact the effectiveness of self-talk interventions. Therefore, we examined how the effectiveness of first-person and second-person pronoun framed strategic self-talk influences endurance performance, with a central interest on the interaction between participants’ performance-oriented trait of narcissism and pronoun framed self-talk. Mechanistic variables (i.e., effort perception and motivational interpretation of self-talk) were also considered to provide insight into why any interactive effects might have emerged. Participants were 40 physically active males (Mage = 19.32, SD = 1.20). To assess performance, the participants completed a 10km cycle time trial at baseline (perceived effort was assessed at 1km intervals). The participants were classified into either higher or lower narcissism experimental groups and randomly assigned to use either first-person or second-person self-talk during a subsequent post-test time trial. ANOVAs indicated that participants in the higher-narcissism first-person self-talk group as well as those in the lower-narcissism second-person group exhibited comparatively better endurance performance. Although no significant differences were found for effort perception, a similarly shaped interaction emerged for participants’ motivational interpretation of their self-talk. Together our findings support the presence of a narcissism × pronoun self-talk interaction, at least for endurance tasks, which challenges the “one size fits all” approach to self-talk interventions present in the existing literature.