While emotional intelligence (EI) is recognized as a resource in social interactions, we hypothesized a positive association with stress in socially evaluative contexts. In particular, we expected emotion recognition, the core component of EI, to inflict stress on individuals in negatively valenced interactions. We expected this association to be stronger for status-driven individuals, that is, for individuals scoring high on basal testosterone. In a laboratory experiment, N = 166 male participants underwent the Trier Social Stress Test (Kirschbaum, Pirke, & Hellhammer, 1993). As expected, EI measured by the Mayer–Salovey–Caruso Emotional Intelligence Test (MSCEIT V2.0; Mayer et al., 2003) predicted higher cortisol reactivity, including slower recovery from stress. The effect was moderated by basal testosterone, such that the association was positive when basal testosterone was high but not when it was low. On the component level of EI, the interaction was replicated for negative emotion recognition. These findings lend support to the hypothesis that EI is associated with higher activity of the hypothalamic-pituitary-adrenal axis in contexts where social status is at stake, particularly for those individuals who are more status-driven. Thus, the effects of EI are not unequivocally positive: While EI may positively affect the course of social interactions, it also inflicts stress on the emotionally intelligent individuals themselves.