Emotional intelligence has become a very popular topic in organizational research (Joseph & Newman, 2010; Mayer, Roberts, & Barsade, 2008), partly as a response to contentions that emotional intelligence predicts job performance as robustly as cognitive ability does (Goleman, 1995). The majority of previous research on the relationship between emotional intelligence and job performance has examined emotional intelligence as an individual difference construct that acts as a direct determinant of job performance (e.g. Carmeli & Josman, 2006). However, research has suggested job-relevant knowledge and skill are direct determinants of job performance and that individual differences in abilities and traits are antecedents of job knowledge (Campbell, Gasser, & Oswald, 1996; Motowidlo, Borman, & Schmit, 1997). Consequently, according to this rationale, emotional intelligence may only affect job performance through its effect on knowledge. This investigation examined whether prosocial knowledge mediates the relationship between emotional intelligence and prosocial skill in role-play simulations of service encounters in medicine using a sample of 199 undergraduate students. Secondary purposes were to replicate results from earlier work demonstrating personality traits affect skill primarily through their effects on knowledge and to explore the construct and predictive validity of job knowledge further. Individual tests of hypotheses were conducted and the overall pattern of relations among study variables is summarized by a path analytic model. Analyses revealed that prosocial knowledge measured by a single-response situational judgment test mediated the effects of agreeableness, conscientiousness, and emotion management on prosocial skill displayed in role-play simulations. Emotion understanding was causally related to emotion management as theorized by hierarchical models of emotional intelligence (Joseph & Newman, 2010), but unexpectedly, emotional stability was not. Results clarify the role of emotional intelligence as a distal antecedent of job performance rather than a more proximal performance determinant.